Human Clinical Studies of Herbal Teas: What Exists and Limits
The human trials of herbal tea we could read, what each tested and found, and the limits that keep most results from guiding everyday choices.
Content last updated . This page has not been medically reviewed.
Safety first. Natural products can cause side effects and interact with medicines. Read herbal tea and extract safety before you use anything on this page.
Do trials in people show that herbal tea does anything? A few do, for a few teas, in small groups. This page lists the human studies of the drink itself that we read, says what each can and cannot tell you, and separates them from studies of capsules and extracts. It is information only.
What counts here
A human clinical study gives a tea, or a comparison, to people and measures what happens. We only list studies where the tea was the thing tested. A randomized trial divides people by chance into groups, which makes the comparison fairer. [1] Studies in cells or animals are covered on human studies versus lab research.
Trials of single teas we read
| Tea | Who and how long | What the source found | Strength of evidence |
|---|---|---|---|
| Hibiscus (Hibiscus sabdariffa L. [2]) | 65 adults with pre- or mild high blood pressure, 6 weeks, placebo-controlled, three 240 mL servings a day [1] | Systolic pressure fell more than with placebo (-7.2 vs -1.3 mm Hg); diastolic did not differ from placebo [1] | Promising but limited |
| Chamomile (Matricaria chamomilla L.; M. recutita is a synonym [2]) | 64 adults with type 2 diabetes, 8 weeks, tea (3 g in 150 mL, three times a day) versus water, single-blind [3] | Lower HbA1c, insulin and cholesterol markers than the water group [3] | Promising but limited; a second paper reports the same design and group size and may describe the same trial [4] |
| Chamomile | 80 postnatal women with poor sleep, tea for 2 weeks versus usual care [5] | Better sleep-related scores and depression symptoms right after, but similar scores at 4 weeks [5] | Mixed; not placebo-controlled |
| Chamomile, effect on clotting tests | 8 healthy volunteers, one tea dose versus capsule [6]; 12 healthy volunteers, 7 days of tea versus capsule versus placebo [7] | The one-dose trial found a small rise in prothrombin time with tea; the 7-day trial found no rise [6][7] | Mixed, very small; clinical meaning unclear |
| Peppermint | 25 healthy adults, 200 mL tea versus placebo, single-blind [8] | Better scores on a computer cognition battery and larger changes in a near-infrared blood-flow measure in the tea group [8] | Promising but limited; one short test in healthy people |
| Ginger | Not applicable | Proper evidence not found: no trial of ginger tea tagged as randomized in our PubMed search of "ginger tea" [9] | Insufficient evidence |
Measures such as HbA1c, cholesterol and clotting tests are lab markers, not whether people felt or lived better. NIH adds that many ginger studies tested supplements rather than foods. [10]
Reviews that pool teas
- A 2024 meta-analysis of 6 randomized trials (235 women with polycystic ovary syndrome) found herbal tea linked with lower weight, BMI and fasting glucose than control. Its authors call for larger trials. [11]
- A 2022 review of 16 studies (832 people with type 2 diabetes) found moderate effects of camellia and herbal teas on several markers, and said more high-quality trials are needed. [12]
- A 2019 review of 6 trials of Chinese herbal tea for chronic pharyngitis rated their quality low and drew no credible conclusion on effect. [13]
- A scoping review of rooibos tea in people found 18 publications and 488 participants, and called the evidence limited. [14]
Pooled reviews mix different plants and recipes, so a result does not belong to any one tea.
Limits that apply to nearly all of it
- Small and short. Most trials above had under 100 people and ran weeks, not years. [4][8]
- Weak comparison groups. Water or usual care is not a placebo tea, so people know what they drink. [3][5]
- Different preparations. The trials above used different amounts and servings, so results do not combine neatly. [1][3][8]
- Healthy volunteers. Results in healthy adults may not carry to people with illness. [8]
Frequently asked questions
Is there a human trial of hibiscus tea?
Yes. A randomized, double-blind, placebo-controlled trial in 65 adults found a drop in systolic blood pressure over 6 weeks. [1]
Do trials show chamomile tea helps sleep?
One trial in postnatal women found short-term gains that were gone at 4 weeks, and NIH says there is very little information on chamomile and insomnia. [5][16]
Why are there so few trials of tea?
The sources we read do not say why. They show small samples and few tea-specific trials. [14]
Do meta-analyses of herbal tea prove it works?
No. They pool different plants and say larger, better trials are needed. [11][12]
Should I tell my doctor about herbal tea?
Yes, especially with medicines or before surgery. [16]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
Safety and who should ask first
A trial finding is not a recommendation. Hibiscus lowered blood pressure in trials, so ask a pharmacist or clinician if you take blood-pressure medicine. [15] NIH says chamomile may interact with warfarin and liver-processed drugs, and little is known about use in pregnancy or breastfeeding. [16] Tell your surgical team about any herbal tea before an operation; see supplements before surgery and how interactions work.
For single teas see chamomile tea, hibiscus tea and peppermint tea.