Stevia (steviol glycosides)

Also called steviol glycosides, stevioside, rebaudioside A, Reb A, Stevia rebaudiana, E960.

Best grade C across 18 graded outcomes Botanical Last checked Oct 21, 2026

The short answer

From 28 sources

Stevia sweeteners are purified steviol glycosides with almost no calories. At sweetener amounts they did not change glucose or blood pressure in trials. Two Chinese trials found lower blood pressure with 750 to 1,500 mg of stevioside a day (grade C), at or above the acceptable daily intake for a 70 kg adult; pooled analyses show small, uncertain effects.

Every graded outcome on this page

18 rows · line = participants

Grade CBlood glucose, insulin and HbA1c (pooled)Blood sugarn 1,439
Grade CFasting glucose and HbA1c in adults (pooled)Blood sugarn 871
Grade CBlood pressure, fasting glucose and lipids (pooled)Blood pressuren 756
Grade CDiabetes markers, blood pressure and lipids (pooled)Blood sugarn 462
Grade CAppetite scores (pooled)Appetiten 428
Grade CBlood pressure in mild hypertension (2 years)Blood pressuren 168
Grade CBody weight with sweetened drinks for 12 weeksWeightn 123
Grade CGlucose tolerance and gut microbes after 2 weeks of sweetener sachetsGut bacterian 120
Grade CBlood pressure in hypertension (1 year)Blood pressuren 106
Grade CGut microbiome and short-chain fatty acids vs a sucrose drinkGut bacterian 59
Grade DGlucose regulation and cardiometabolic markers after 14 days of sweetened drinksBlood sugarn 39
Grade DGlucose tolerance in type 2 diabetes the day after 3 gBlood sugarn 30
Grade DDaily energy intake and glucose after a sweetened drinkAppetiten 30
Grade DGlucose tolerance and weight with a daily stevia extractBlood sugarn 28
Grade DAppetite and total energy intake at lunch after a stevia drinkAppetiten 20
Grade DAfter-meal glucose in type 2 diabetes (one dose)Blood sugarn 12
No effectHbA1c and fasting glucose in type 2 diabetesBlood sugarn 122
No effectBlood pressure in people with normal or low-normal pressureBlood pressuren 100
15 more graded outcomes, in the evidence chapter

Key facts

What it is: Sweet glycosides (stevioside, rebaudioside A and others) from Stevia rebaudiana leaves, 200 to 400 times sweeter than table sugar^1,2
US status: High-purity (95% minimum) steviol glycosides are used under GRAS notices FDA did not question; stevia leaf and crude extracts are not considered GRAS as sweeteners¹
Safe intake: Acceptable daily intake 4 mg per kg of body weight as steviol equivalents (JECFA), which FDA gives as 12 mg/kg for rebaudioside A¹
How the body handles it: Not digested or absorbed in the upper gut; gut bacteria release steviol, which is absorbed, turned into steviol glucuronide and excreted mainly in urine^1,14,13
Blood pressure: Two Chinese trials (106 and 174 people) found lower blood pressure with 750 to 1,500 mg a day of stevioside (C); rebaudioside A at 1,000 mg a day and a crude stevioside extract did not beat placebo, and shorter trials were negative^5,4,6,26,25
Blood sugar: 1,000 mg a day of rebaudioside A for 16 weeks did not change HbA1c in type 2 diabetes (no effect); pooled trials show a small fall in fasting glucose of low certainty^11,9,10
Weight: In a 12-week trial, drinks sweetened with rebaudioside A did not change weight, while sucrose drinks added 1.85 kg¹⁵
Gut bacteria: Four weeks of a stevia drink did not change the gut microbiome compared with a sucrose drink; a 2-week sachet study found stevia altered gut microbes without impairing glucose tolerance^16,17

What it is

Stevia sweeteners are purified steviol glycosides from the leaves of Stevia rebaudiana, 200 to 400 times sweeter than sugar.

How it works

Established

Steviol glycosides are diterpene glycosides that taste intensely sweet. In people they are not hydrolysed by digestive enzymes or absorbed in the upper gut; colon bacteria remove the sugars to release steviol, which is absorbed, conjugated to steviol glucuronide and excreted mainly in urine, while steviol itself appears in faeces. Rebaudioside A and stevioside follow the same pathway, with somewhat lower exposure from rebaudioside A. Glucose-lowering effects seen in mice have not been reproduced with rebaudioside A in people with type 2 diabetes; the mechanism behind the blood-pressure fall in the stevioside trials is not established.^1,13,14

Forms side by side: purified sweeteners, research capsules and leaf

4 forms named in our sources. The rest show what our data holds, and say plainly where it holds nothing yet.

Stevia (steviol glycosides) forms compared

4 forms

Stevia (steviol glycosides) forms compared: 4 forms by 4 attributes. Empty cells are labelled No data yet, or Not graded in the evidence row.
4 forms	High-purity steviol glycosides (tabletop packets, drinks, foods)	Stevioside capsules (research doses)	Rebaudioside A (single purified glycoside)	Stevia leaf and crude extracts
Standardized toMarker compounds, share of extract weight	95%¹At least 95% steviol glycosides (rebaudioside A, stevioside, rebaudioside D or mixtures), the purity in the GRAS notices FDA did not question	Not applicable (stevioside, C₃₈H₆₀O₁₈; about 40% steviol by weight)^2,3	Not applicable (rebaudioside A, C₄₄H₇₀O₂₃; about 33% steviol by weight)^24,3	Not applicable (whole leaf or unpurified extract; glycoside content varies)¹
Absorption in people	Not absorbed in the upper gut; gut bacteria release steviol, which is absorbed and excreted as steviol glucuronide.^1,14	Steviol glucuronide peaks about 8 hours after a dose and has a half-life of about 14 hours.¹⁴	Steviol glucuronide exposure about 10% to 22% lower than after stevioside; peak about 12 hours after a dose.¹⁴	No head-to-head human data.¹
GI tolerance	No adverse effects on gut bacteria, glucose or lipids in a 4-week trial at 25% of the acceptable daily intake.¹⁶	Adverse events did not differ from placebo over 2 years at 1,500 mg a day; reviews list abdominal fullness, stomach pain and dizziness.^5,7	1,000 mg a day for 16 weeks was well tolerated in type 2 diabetes, with no excess of low blood sugar.¹¹	FDA does not consider stevia leaf or crude extracts GRAS, and their import as sweeteners is not permitted.¹
Studied for	Weight, appetite and gut-bacteria studies (C to D)	Blood pressure in hypertension (C, at 750 to 1,500 mg/day)	HbA1c in type 2 diabetes (no effect); blood pressure in normal pressure (no effect)	Not graded here (a crude stevioside trial up to 15 mg/kg/day found no blood-pressure effect beyond placebo)

High-purity steviol glycosides (tabletop packets, drinks, foods)
Standardized to
95%¹At least 95% steviol glycosides (rebaudioside A, stevioside, rebaudioside D or mixtures), the purity in the GRAS notices FDA did not question
Absorption in people
Not absorbed in the upper gut; gut bacteria release steviol, which is absorbed and excreted as steviol glucuronide.^1,14
GI tolerance
No adverse effects on gut bacteria, glucose or lipids in a 4-week trial at 25% of the acceptable daily intake.¹⁶
Studied for
Weight, appetite and gut-bacteria studies (C to D)
Stevioside capsules (research doses)
Standardized to
Not applicable (stevioside, C₃₈H₆₀O₁₈; about 40% steviol by weight)^2,3
Absorption in people
Steviol glucuronide peaks about 8 hours after a dose and has a half-life of about 14 hours.¹⁴
GI tolerance
Adverse events did not differ from placebo over 2 years at 1,500 mg a day; reviews list abdominal fullness, stomach pain and dizziness.^5,7
Studied for
Blood pressure in hypertension (C, at 750 to 1,500 mg/day)
Rebaudioside A (single purified glycoside)
Standardized to
Not applicable (rebaudioside A, C₄₄H₇₀O₂₃; about 33% steviol by weight)^24,3
Absorption in people
Steviol glucuronide exposure about 10% to 22% lower than after stevioside; peak about 12 hours after a dose.¹⁴
GI tolerance
1,000 mg a day for 16 weeks was well tolerated in type 2 diabetes, with no excess of low blood sugar.¹¹
Studied for
HbA1c in type 2 diabetes (no effect); blood pressure in normal pressure (no effect)
Stevia leaf and crude extracts
Standardized to
Not applicable (whole leaf or unpurified extract; glycoside content varies)¹
Absorption in people
No head-to-head human data.¹
GI tolerance
FDA does not consider stevia leaf or crude extracts GRAS, and their import as sweeteners is not permitted.¹
Studied for
Not graded here (a crude stevioside trial up to 15 mg/kg/day found no blood-pressure effect beyond placebo)

What the human research says

Human trials only, one row per outcome. The form line shows what each trial actually tested. Grades follow our evidence-grading policy.

Trial map

18 graded outcomes, largest study first

18 graded outcomes: 4 improved, 5 mixed, 9 no effect. The largest, a 2024 meta-analysis of 1,439 people on blood glucose, insulin and HbA1c (pooled), found mixed results.

By sizeTable

1Blood glucose, insulin and HbA1c (pooled)

Mixed · Grade C · Meta-analysis, 1,439 people, 2024

Participants of 26 studies

Dose, result and source

Every graded outcome on this page, largest study first
Outcome	Found	People	Study	Who was studied
Blood glucose, insulin and HbA1c (pooled)	Mixed	1,439	Grade C · Meta-analysis, 2024	Participants of 26 studies
Fasting glucose and HbA1c in adults (pooled)	Mixed	871	Grade C · Meta-analysis, 2024	Adults in 12 randomized trials (48% women)
Blood pressure, fasting glucose and lipids (pooled)	Mixed	756	Grade C · Meta-analysis, 2015	Participants of 9 randomized trials, most at high cardiovascular risk
Diabetes markers, blood pressure and lipids (pooled)	Mixed	462	Grade C · Meta-analysis, 2019	Participants of 9 randomized trials (7 studies)
Appetite scores (pooled)	No effect	428	Grade C · Meta-analysis, 2025	Adults in 11 clinical trials
Blood pressure in mild hypertension (2 years)	Improved	168	Grade C · RCT, 2003	Chinese adults aged 20 to 75 with mild essential hypertension (174 enrolled, 87 men and 87 women)
Body weight with sweetened drinks for 12 weeks	No effect	123	Grade C · RCT, 2019	Adults aged 18 to 60 with overweight or obesity (154 randomized to five drinks)
HbA1c and fasting glucose in type 2 diabetes	No effect	122	No-effect grade · RCT, 2008	Men and women aged 33 to 75 with type 2 diabetes
Glucose tolerance and gut microbes after 2 weeks of sweetener sachets	Mixed	120	Grade C · RCT, 2022	Healthy adults randomized to saccharin, sucralose, aspartame, stevia, glucose vehicle or no supplement
Blood pressure in hypertension (1 year)	Improved	106	Grade C · RCT, 2000	Chinese adults aged 28 to 75 with diastolic pressure 95 to 110 mm Hg
Blood pressure in people with normal or low-normal pressure	No effect	100	No-effect grade · RCT, 2008	Healthy adults aged 18 to 73, mostly women, with resting pressure around 110/70 mm Hg
Gut microbiome and short-chain fatty acids vs a sucrose drink	No effect	59	Grade C · RCT, 2024	Healthy adults, mean age 31 (36 women, 23 men)
Glucose regulation and cardiometabolic markers after 14 days of sweetened drinks	No effect	39	Grade D · Crossover RCT, 2026	Healthy, normal-weight adults (Sweden)
Glucose tolerance in type 2 diabetes the day after 3 g	No effect	30	Grade D · Crossover RCT, 2022	Adults with type 2 diabetes on metformin or no treatment, University Hospitals Leuven (Belgium), 86.7% men
Daily energy intake and glucose after a sweetened drink	No effect	30	Grade D · Crossover RCT, 2017	Healthy men
Glucose tolerance and weight with a daily stevia extract	No effect	28	Grade D · Open-label trial, 2020	Healthy normal-weight adults, mean age 25
Appetite and total energy intake at lunch after a stevia drink	Improved	20	Grade D · Crossover RCT, 2020	Healthy adults, mean age 27, 55% women
After-meal glucose in type 2 diabetes (one dose)	Improved	12	Grade D · Crossover RCT, 2004	People with type 2 diabetes

Every graded row

18 rows

Form tested

Showing 18 of 18 outcomes. Human trials only.

Grade CBlood glucose, insulin and HbA1c (pooled)Blood sugar−3.84 mg/dL glucoseMeta-analysis · 26 studies
n 1,439 · 2024

Mixed Blood glucose fell slightly (−3.84 mg/dL, 95% CI −7.15 to −0.53; low certainty), mainly in people with higher BMI, diabetes or hypertension. Insulin and HbA1c did not change significantly.

Design: Meta-analysis, 26 studies, n = 1,439, 2024
Population: Participants of 26 studies
Form: Stevia
Dose: As given in each study; glucose fell at 3,342 mg/day or more in dose-response analysis
Duration: Varied by study
Source: Reference 10 · PubMed 39098209 (opens PubMed)

Grade CFasting glucose and HbA1c in adults (pooled)Blood sugarFBG −4.10 mg/dL; HbA1c no changeMeta-analysis · 12 studies
n 871 · 2024

Mixed Fasting glucose was 4.10 mg/dL lower than control (95% CI −6.55 to −1.65); HbA1c did not differ (0.01%). The overall quality of evidence was rated low.

Design: Meta-analysis, 12 studies, n = 871, 2024
Population: Adults in 12 randomized trials (48% women)
Form: Steviol glycosides
Dose: As given in each included trial
Duration: Varied by trial
Source: Reference 9 · PubMed 38512280 (opens PubMed)

Grade CBlood pressure, fasting glucose and lipids (pooled)Blood pressureSBP −2.98 mm Hg (n.s.)Meta-analysis · 9 studies
n 756 · 2015

Mixed Systolic pressure did not differ significantly from placebo (−2.98 mm Hg, 95% CI −6.23 to 0.27); diastolic pressure and fasting glucose fell significantly, with no effect on lipids. Rebaudioside A had no significant effect on blood pressure or other risk factors. Heterogeneity was significant.

Design: Meta-analysis, 9 studies, n = 756, 2015
Population: Participants of 9 randomized trials, most at high cardiovascular risk
Form: Stevioside or rebaudioside A
Dose: As given in each included trial
Duration: Varied by trial
Source: Reference 7 · PubMed 25412840 (opens PubMed)

Grade CDiabetes markers, blood pressure and lipids (pooled)Blood sugarNo effect on HbA1cMeta-analysis · 9 studies
n 462 · 2019

Mixed The review reports a significant fall in systolic pressure (−6.32 mm Hg) and non-significant changes in BMI, diastolic pressure, fasting glucose and cholesterol, with no effect on HbA1c. Heterogeneity was significant.

Design: Meta-analysis, 9 studies, n = 462, 2019
Population: Participants of 9 randomized trials (7 studies)
Form: Steviol glycosides
Dose: As given in each included trial
Duration: Varied by trial
Source: Reference 8 · PubMed 31438580 (opens PubMed)

Grade CAppetite scores (pooled)AppetiteNo change in appetiteMeta-analysis · 11 studies
n 428 · 2025

No effect Stevia had no significant effect on any appetite score (very low to moderate certainty); too few trials reported energy intake to pool.

Design: Meta-analysis, 11 studies, n = 428, 2025
Population: Adults in 11 clinical trials
Form: Stevia
Dose: As given in each trial
Duration: Varied by trial
Source: Reference 20 · PubMed 39887829 (opens PubMed)

Grade CBlood pressure in mild hypertension (2 years)Blood pressure150/95 to 140/89 mm HgRCT
n 168 · 2003

Improved Systolic pressure fell from 150 to 140 mm Hg and diastolic from 95 to 89 mm Hg with stevioside, significantly more than placebo; the effect began about a week after starting. Body mass index, blood tests and adverse events did not differ from placebo.

Design: RCT, n = 168, 2003
Population: Chinese adults aged 20 to 75 with mild essential hypertension (174 enrolled, 87 men and 87 women)
Form: Stevioside powder capsules vs placebo
Dose: 500 mg three times a day (1,500 mg/day)
Duration: 2 years
Source: Reference 5 · PubMed 14693305 (opens PubMed)

Grade CBody weight with sweetened drinks for 12 weeksWeightNo weight gain (sucrose +1.85 kg)RCT
n 123 · 2019

No effect Weight rose with sucrose (+1.85 kg) and saccharin (+1.18 kg) but did not change significantly with rebaudioside A. Glucose tolerance did not change with any sweetener.

Design: RCT, n = 123, 2019
Population: Adults aged 18 to 60 with overweight or obesity (154 randomized to five drinks)
Form: Drinks sweetened with rebaudioside A, aspartame, saccharin, sucralose or sucrose
Dose: 1.25 to 1.75 L of sweetened drink a day
Duration: 12 weeks
Source: Reference 15 · PubMed 30997499 (opens PubMed)

Grade CGlucose tolerance and gut microbes after 2 weeks of sweetener sachetsGut bacteriaMicrobes changed; glucose not impairedRCT
n 120 · 2022

Mixed Each sweetener, stevia included, changed stool and oral microbes and plasma metabolites. Saccharin and sucralose significantly impaired glucose responses; stevia was not reported to.

Design: RCT, n = 120, 2022
Population: Healthy adults randomized to saccharin, sucralose, aspartame, stevia, glucose vehicle or no supplement
Form: Stevia sachets (below the acceptable daily intake)
Dose: Sachets daily, below the acceptable daily intake
Duration: 2 weeks
Source: Reference 17 · PubMed 35987213 (opens PubMed)

Grade CBlood pressure in hypertension (1 year)Blood pressure166/105 to 153/90 mm HgRCT
n 106 · 2000

Improved After 3 months, systolic pressure fell from 166.0 to 152.6 mm Hg and diastolic from 104.7 to 90.3 mm Hg, and the effect persisted for the year. Lipids and glucose did not change and no significant adverse effects were seen.

Design: RCT, n = 106, 2000
Population: Chinese adults aged 28 to 75 with diastolic pressure 95 to 110 mm Hg
Form: Stevioside capsules vs placebo
Dose: 250 mg three times a day (750 mg/day)
Duration: 1 year
Source: Reference 4 · PubMed 10971305 (opens PubMed)

Grade CGut microbiome and short-chain fatty acids vs a sucrose drinkGut bacteriaNo microbiome changeRCT
n 59 · 2024

No effect No differences in the gut microbiome at any level or in faecal short-chain fatty acids; fasting glucose, insulin, lipids and blood pressure did not differ. BMI rose 0.3 kg/m² with sucrose compared with stevia.

Design: RCT, n = 59, 2024
Population: Healthy adults, mean age 31 (36 women, 23 men)
Form: Drink with 25% of the acceptable daily intake of stevia vs drink with 30 g sucrose
Dose: 16 oz drink daily
Duration: 4 weeks, then 4-week washout
Source: Reference 16 · PubMed 38408729 (opens PubMed)

Grade DGlucose regulation and cardiometabolic markers after 14 days of sweetened drinksBlood sugarNo difference vs sucrose or saccharinCrossover RCT
n 39 · 2026

No effect Glucose regulation did not differ between steviol glycosides, saccharin (220 mg/day) and sucrose (66 g/day). Fasting insulin rose from baseline after stevia, and all three drinks raised after-meal insulin. Triglycerides were higher after sucrose than saccharin; cholesterol, CRP and appetite did not differ.

Design: Crossover RCT, n = 39, 2026
Population: Healthy, normal-weight adults (Sweden)
Form: Steviol glycosides in drinks
Dose: 220 mg/day
Duration: 14 days per drink, with washouts
Source: Reference 27 · PubMed 42514406 (opens PubMed)

Grade DGlucose tolerance in type 2 diabetes the day after 3 gBlood sugarNo change in glucose curveCrossover RCT
n 30 · 2022

No effect Rebaudioside A was absorbed and converted to steviol and steviol glucuronide, but did not lower the 2-hour glucose curve (−0.7 h·mg/dL, 95% CI −22.3 to 20.9) or change insulin or C-peptide.

Design: Crossover RCT, n = 30, 2022
Population: Adults with type 2 diabetes on metformin or no treatment, University Hospitals Leuven (Belgium), 86.7% men
Form: Rebaudioside A vs placebo (open-label crossover)
Dose: 3 g single dose
Duration: Glucose tolerance test 19 hours later
Source: Reference 13 · PubMed 36057030 (opens PubMed)

Grade DDaily energy intake and glucose after a sweetened drinkAppetiteNo difference in daily energyCrossover RCT
n 30 · 2017

No effect People ate more at lunch after the no-calorie drinks, so the energy saved was fully made up and total daily intake did not differ. Total 3-hour glucose and insulin areas did not differ between the four drinks.

Design: Crossover RCT, n = 30, 2017
Population: Healthy men
Form: Stevia-, monk fruit-, aspartame- or sucrose-sweetened drink
Dose: One drink mid-morning
Duration: Single test days
Source: Reference 21 · PubMed 27956737 (opens PubMed)

Grade DGlucose tolerance and weight with a daily stevia extractBlood sugarNo glucose change; −0.22 vs +0.89 kgOpen-label trial
n 28 · 2020

No effect Glucose and insulin responses to a glucose tolerance test did not change. Weight was stable with stevia (−0.22 kg) while the control group gained 0.89 kg, and energy intake fell in the stevia group.

Design: Open-label trial, n = 28, 2020
Population: Healthy normal-weight adults, mean age 25
Form: Daily stevia extract vs no stevia (randomized, open-label)
Dose: A stevia extract daily (amount not stated in the abstract)
Duration: 12 weeks
Source: Reference 18 · PubMed 33036155 (opens PubMed)

Grade DAppetite and total energy intake at lunch after a stevia drinkAppetite727 vs 832 kcalCrossover RCT
n 20 · 2020

Improved Total energy (drink plus lunch) was lower after the stevia drink (727 kcal) than after water (832 kcal, P = 0.013). Blood glucose after stevia was similar to water and lower than after the caloric drinks.

Design: Crossover RCT, n = 20, 2020
Population: Healthy adults, mean age 27, 55% women
Form: 240 ppm stevia drink vs water, glucose, sucrose or maltodextrin drinks (330 mL)
Dose: One 330 mL drink before lunch
Duration: Single test visits
Source: Reference 19 · PubMed 32125421 (opens PubMed)

Grade DAfter-meal glucose in type 2 diabetes (one dose)Blood sugar−18% glucose areaCrossover RCT
n 12 · 2004

Improved Stevioside cut the incremental glucose area under the curve by 18% (P = 0.013) and raised the insulinogenic index by about 40%; insulin, GLP-1 and GIP areas did not change significantly.

Design: Crossover RCT, n = 12, 2004
Population: People with type 2 diabetes
Form: 1 g stevioside vs 1 g maize starch added to a test meal
Dose: 1 g with one meal
Duration: Single test meals
Source: Reference 12 · PubMed 14681845 (opens PubMed)

No effectHbA1c and fasting glucose in type 2 diabetesBlood sugarHbA1c +0.11% vs +0.09%RCT
n 122 · 2008

No effect HbA1c changed by 0.11% with rebaudioside A and 0.09% with placebo (p = 0.355); fasting glucose, insulin, C-peptide, blood pressure, weight and lipids did not differ. Hypoglycaemic episodes were not more common.

Design: RCT, n = 122, 2008
Population: Men and women aged 33 to 75 with type 2 diabetes
Form: Rebaudioside A vs placebo
Dose: 1,000 mg/day
Duration: 16 weeks
Source: Reference 11 · PubMed 18555575 (opens PubMed)

No effectBlood pressure in people with normal or low-normal pressureBlood pressureNo change vs placeboRCT
n 100 · 2008

No effect Rebaudioside A did not change resting blood pressure, mean arterial pressure, heart rate or 24-hour ambulatory pressure compared with placebo.

Design: RCT, n = 100, 2008
Population: Healthy adults aged 18 to 73, mostly women, with resting pressure around 110/70 mm Hg
Form: Rebaudioside A vs placebo
Dose: 1,000 mg/day
Duration: 4 weeks
Source: Reference 6 · PubMed 18555574 (opens PubMed)

What the numbers add up to

Each point is worked out from the data and the sources numbered on this page, with the arithmetic shown under it. It adds no new data and is not medical advice.

The blood-pressure trials used about twice the daily limit

1,500 mg of stevioside a day is about 593 mg of steviol equivalents, or 8.5 mg per kg for a 70 kg adult: roughly twice the acceptable daily intake of 4 mg/kg.^5,1,2,3

The daily limit in rebaudioside A

For a 70 kg adult, FDA's 12 mg/kg for rebaudioside A is 840 mg a day; the 16-week diabetes trial used 1,000 mg a day and found no change in HbA1c.^1,11

Sweetener use vs research use

The glucose fall in the 2024 dose-response analysis appeared at 3,342 mg of stevia a day or more, which is well above what sweetening food and drinks requires.¹⁰

Studied doses

Doses used in research, shown against the official limits. They describe the studies, not a recommendation for you.

Studied doses in research, not advice. These are amounts used in research, not personal advice. Your clinician or pharmacist can say what fits you.

Doses studied in human trials

Doses studied in human trials ranged from 250 mg of stevioside three times a day (750 mg/day) to 500 mg three times a day (1,500 mg/day) for 1 to 2 years in blood-pressure trials, 1,000 mg/day of rebaudioside A for 4 to 16 weeks, and single test doses of 1 g of stevioside with a meal and 3 g of rebaudioside A. Weight, appetite and microbiome studies used sweetener amounts in drinks and sachets. These are research amounts, not product directions.^{4,5,6,11,12,13,15,16}

When trials measured an effect

3 outcomes

How long trials ran before effects were measured
Outcome	Measured at	Basis
Blood pressure (stevioside, hypertension)	From about 1 week on home readings; significant by 3 months and sustained to 1 and 2 years	Two Chinese placebo-controlled trials^5,4
HbA1c (rebaudioside A, type 2 diabetes)	No change at 16 weeks	Placebo-controlled trial in 122 people¹¹
Blood steviol glucuronide after a dose	Peaks about 8 to 12 hours after a dose; half-life about 14 hours	Pharmacokinetic study in healthy men¹⁴

Interactions and safety

Published interactions only: each row cites a source. No row means none was found in our sources, not that a combination is safe.

What Stevia meets

2 cited rows

Medicines 2

Stevia2 medicines · 0 supplements

Supplements and foods 0

No supplement rows in our sources.

Reference summary

Interactions

2 agents

Published interactions, grouped by kind
Agent	What the source says	Evidence
Medicines2
Blood-pressure medicinese.g. Verapamil	MSKCC: animal studies suggest stevia may add to the blood-pressure-lowering effect; clinical relevance has not been determined. Stevioside at 750 to 1,500 mg a day lowered blood pressure in two trials, far above sweetener amounts.^28,4,5	Review
Glucose-lowering medicines	MSKCC: animal studies suggest stevia may affect blood glucose; clinical relevance has not been determined. 1,000 mg a day of rebaudioside A did not change HbA1c in type 2 diabetes.^28,11	Review

Check with what you take

Kept on this device only

Add what you take. We list every interaction our evidence profiles cite between Stevia and your picks, and between your picks.

Add a supplement or medicine

Most cited:

Add a supplement or medicine you take to check it against Stevia.

Starting from one medicine? The interaction checker rates how serious each pair is, on a curated list of 19 entries for 11 medicines. This cabinet has no severity rating: it lists every interaction row our supplement pages cite.

Who should be careful

3 groups

Pregnancy and breastfeeding: Pregnancy: MSKCC advises against stevia in pregnancy because little is known about how non-nutritive sweeteners affect offspring. Breastfeeding: NIH's LactMed found no data on stevia in breast milk or in nursing infants; it notes the risk to a breastfed infant appears low and that a sweetener with more data may be preferred while nursing a newborn or preterm infant.^28,22
People on blood-pressure medicines: Stevioside at 750 to 1,500 mg a day lowered blood pressure in two trials. Sweetener amounts are far lower, but anyone using stevia extracts at research doses alongside blood-pressure treatment should tell their clinician.^4,5
Leaf and crude extracts: FDA does not consider whole-leaf or crude stevia extracts GRAS as sweeteners; the safety data behind the acceptable daily intake are for purified steviol glycosides.¹

How to read the label

What to check on the Supplement Facts panel.

How to read the label

3 checks

Which glycoside, and how pure

US stevia sweeteners use high-purity (95% minimum) steviol glycosides such as rebaudioside A, stevioside or rebaudioside D. Leaf and crude extracts are a different, non-GRAS category.¹

What else is in the packet

Stevia is 200 to 400 times sweeter than sugar, so little is needed for sweetness. Tabletop products may add other sweeteners, such as erythritol, that have their own tolerance and safety data; check the ingredient list.^1,23

Brand names

FDA lists Truvia, PureVia and Enliten as examples of steviol glycoside sweeteners.¹

Have the bottle? Photograph its Supplement Facts panel and see each line, stevia (steviol glycosides) included, beside the doses and forms the trials used in Label X-Ray.

Frequently asked questions

Short answers to common questions; the numbers link to the sources below.

Common questions

8 answers

Does stevia raise blood sugar?

No, in the trials we found. 1,000 mg a day of rebaudioside A for 16 weeks did not change HbA1c in type 2 diabetes, and a stevia drink before lunch left blood glucose similar to water.^11,19

Does stevia lower blood pressure?

Two Chinese trials found lower blood pressure with 750 to 1,500 mg of stevioside a day for 1 to 2 years; the higher dose is about twice the acceptable daily intake. Rebaudioside A and a crude stevioside extract did not beat placebo, and a 2010 review found shorter 1- to 3-month studies negative. Grade C.^4,5,6,26,25

Can stevia help with weight loss?

Swapping sugar for stevia avoids the calories. In a 12-week trial, sucrose drinks added 1.85 kg while rebaudioside A drinks did not change weight; a stevia drink before lunch lowered total energy intake in one study, though another found the saved calories were made up later in the day.^15,19,21

Is stevia safe?

High-purity steviol glycosides are used in the US under GRAS notices FDA did not question. The acceptable daily intake is 4 mg per kg of body weight as steviol equivalents (12 mg/kg as rebaudioside A, about 840 mg a day for a 70 kg adult), far above normal use. Trials found no change in glucose, blood pressure or gut bacteria at sweetener amounts.^1,16,11

Does stevia harm gut bacteria?

A 4-week trial found no change in the gut microbiome compared with a sucrose drink. A 2-week study found stevia changed gut microbes, but unlike saccharin and sucralose it was not reported to impair glucose tolerance.^16,17

Is stevia safe in pregnancy or while breastfeeding?

MSKCC advises against stevia in pregnancy because effects of non-nutritive sweeteners on offspring are not known. For breastfeeding, LactMed found no milk data; it rates the risk to a breastfed infant as appearing low but suggests a sweetener with more data for a newborn or preterm infant.^28,22

Is stevia better than artificial sweeteners?

In head-to-head trials, rebaudioside A and aspartame did not cause the weight gain seen with sucrose and saccharin, and stevia did not impair glucose tolerance in a study where saccharin and sucralose did. That is limited evidence, not a ranking.^15,17

What is the difference between stevia leaf and stevia sweeteners?

Sweeteners use purified steviol glycosides (95% minimum purity in the US). FDA does not consider the leaf or crude extracts GRAS as sweeteners and does not permit their import for that use.¹

Sources

28 sources: trial abstracts, NIH fact sheets and reference databases. Numbers in the text link here.

Sources cited on this page

28 sources

1Aspartame and Other Sweeteners in Food (steviol glycosides, Luo Han Guo, acceptable daily intakes)Fact sheet U.S. Food and Drug Administration, 2025 · fda.gov
2PubChem compound summary: Stevioside (CID 442089)Database National Library of Medicine, 2026 · pubchem.ncbi.nlm.nih.gov
3PubChem compound summary: Steviol (CID 452967)Database National Library of Medicine, 2026 · pubchem.ncbi.nlm.nih.gov
4A double-blind placebo-controlled study of the effectiveness and tolerability of oral stevioside in human hypertension.RCT Br J Clin Pharmacol, 2000 · PMID 10971305
5Efficacy and tolerability of oral stevioside in patients with mild essential hypertension: a two-year, randomized, placebo-controlled study.RCT Clin Ther, 2003 · PMID 14693305
6The hemodynamic effects of rebaudioside A in healthy adults with normal and low-normal blood pressure.RCT Food Chem Toxicol, 2008 · PMID 18555574
7Effect of the natural sweetener, steviol glycoside, on cardiovascular risk factors: a systematic review and meta-analysis of randomised clinical trials.Meta-analysis Eur J Prev Cardiol, 2015 · PMID 25412840
8Effect of Steviol Glycosides on Human Health with Emphasis on Type 2 Diabetic Biomarkers: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.Meta-analysis Nutrients, 2019 · PMID 31438580
9Effect of steviol glycosides as natural sweeteners on glucose metabolism in adult participants.Meta-analysis Food Funct, 2024 · PMID 38512280
10Effect of stevia on blood glucose and HbA1C: A meta-analysis.Meta-analysis Diabetes Metab Syndr, 2024 · PMID 39098209
11Chronic consumption of rebaudioside A, a steviol glycoside, in men and women with type 2 diabetes mellitus.RCT Food Chem Toxicol, 2008 · PMID 18555575
12Antihyperglycemic effects of stevioside in type 2 diabetic subjects.Crossover RCT Metabolism, 2004 · PMID 14681845
13Pharmacokinetics of Oral Rebaudioside A in Patients with Type 2 Diabetes Mellitus and Its Effects on Glucose Homeostasis: A Placebo-Controlled Crossover Trial.Crossover RCT Eur J Drug Metab Pharmacokinet, 2022 · PMID 36057030
14Pharmacokinetics of rebaudioside A and stevioside after single oral doses in healthy men.PK study Food Chem Toxicol, 2008 · PMID 18555578
15A randomized controlled trial contrasting the effects of 4 low-calorie sweeteners and sucrose on body weight in adults with overweight or obesity.RCT Am J Clin Nutr, 2019 · PMID 30997499
16Comparison of a Daily Steviol Glycoside Beverage compared with a Sucrose Beverage for Four Weeks on Gut Microbiome in Healthy Adults.RCT J Nutr, 2024 · PMID 38408729
17Personalized microbiome-driven effects of non-nutritive sweeteners on human glucose tolerance.RCT Cell, 2022 · PMID 35987213
18Effects of the Daily Consumption of Stevia on Glucose Homeostasis, Body Weight, and Energy Intake: A Randomised Open-Label 12-Week Trial in Healthy Adults.Open-label trial Nutrients, 2020 · PMID 33036155
19Stevia Beverage Consumption prior to Lunch Reduces Appetite and Total Energy Intake without Affecting Glycemia or Attentional Bias to Food Cues: A Double-Blind Randomized Controlled Trial in Healthy Adults.Crossover RCT J Nutr, 2020 · PMID 32125421
20Effects of stevia consumption on appetite in adults: A systematic review and dose-response meta-analysis of randomized controlled trials.Meta-analysis Obes Rev, 2025 · PMID 39887829
21Effects of aspartame-, monk fruit-, stevia- and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake.Crossover RCT Int J Obes (Lond), 2017 · PMID 27956737
22Stevia. Drugs and Lactation Database (LactMed).Database National Institute of Child Health and Human Development, 2024 · PMID 30000908
23A polyol-stevia blended sugar replacer exhibits low glycemic response among human subjects.RCT Clin Nutr ESPEN, 2019 · PMID 31451273
24PubChem compound summary: Rebaudioside A (CID 6918840)Database National Library of Medicine, 2026 · pubchem.ncbi.nlm.nih.gov
25An evidence-based systematic review of stevia by the Natural Standard Research Collaboration.Systematic review Cardiovasc Hematol Agents Med Chem, 2010 · PMID 20370653
26Investigation of the antihypertensive effect of oral crude stevioside in patients with mild essential hypertension.RCT Phytother Res, 2006 · PMID 16775813
27Effects of Short-Term (14-Day) Intake of Sucrose and Non-Caloric Sweeteners on Glucose Regulation, Blood Lipids, Gut Hormones, Inflammation Markers, and Appetite in Healthy Adults: A Randomized Controlled Trial.Crossover RCT Nutrients, 2026 · PMID 42514406
28Stevia (About Herbs)Database Memorial Sloan Kettering Cancer Center, 2022 · mskcc.org

What changed

Every dated change to this page: what changed, who made it and why.

Change log

2 entries

Dated changes to this page
Date	Change	Who and why
Oct 21, 2026evidence, interactions, who should be careful, FAQ	Added a 2026 crossover trial of steviol glycosides vs sucrose and saccharin (PMID 42514406, no difference in glucose regulation). Added blood-pressure and glucose-lowering medicine interactions and a pregnancy note from MSKCC About Herbs (Stevia); both interactions rest on animal data. FAQ trimmed to 8 by folding the daily-limit answer into the safety question; the pregnancy and breastfeeding answer now covers both. Fact-checked against the sources; corrected: the 2026 crossover trial result now notes triglycerides were higher after sucrose than saccharin (lipids did differ), and its grade lowered from C to D as a single small trial.	Writer: Kymata Health EditorialTrigger: A–Z page audit
Oct 19, 2026all	New page. Every PMID checked with NCBI E-utilities (esummary title, year and journal) and its abstract read with efetch, including the LactMed record; FDA's sweetener page (GRAS status, acceptable daily intakes, sweetness) and PubChem molecular weights read on the date shown. The crude-stevioside trial (Ferri 2006) is cited but not graded because its abstract gives no participant count.	Writer: Kymata Health EditorialTrigger: new page

Supplement & Nutrient Encyclopedia. Educational information only; it is not medical advice and does not replace a clinician. Dietary supplements are not intended to diagnose, treat, cure or prevent any disease. Statements about supplements have not been evaluated by the FDA.

Stevia (steviol glycosides)

Medicines 2

Supplements and foods 0

Blood-pressure medicines

Glucose-lowering medicines