Herb Research & Evidence

A lot of published research on an herb does not mean the herb works. It can mean the opposite: many careful studies sometimes end in "not shown to help". This page shows how much research exists for common herbs, what the sources say it found, and how to read herb studies yourself. Research about whether an herb helps is a separate question from whether it is safe, and both questions are only answered for one herb, one form and one use at a time.

Evidence snapshot

The snapshot counts PubMed records that mention each herb. A count shows how much has been published, not whether an herb works. The searches match words in titles and abstracts, so they include laboratory, animal and food research, and curcumin and silymarin are counted with their plants.

Counted 2026-10-22 using NCBI E-utilities. Counts include studies of extracts, blends and non-health topics, and a trial can find no effect. A higher count means more human research has been indexed, not that something works.

Lots of studies, different answers

The table shows what a U.S. government source says about six herbs that are widely studied. The labels in the third column follow this library's evidence labels.

NCCIH also says that some herbal and dietary supplements have failed to show a benefit when scientists studied them. [8] Plain "natural" is not a result: only the studies are.

How to read an herb study

• Which herb, which part, which product? A Cochrane review of echinacea found that the trials used different species and plant parts. [6] NCCIH warns that supplements you buy may differ in important ways from products tested in studies. [9]
• People or the lab? Findings in cells or animals are preclinical. NCCIH says clinical trials give the clearest information about whether a treatment is effective and safe in humans. [10]
• How many people? Small studies are more likely to produce results due to chance. [11]
• Is it one study? A single study rarely provides a final, definitive answer. [12]
• Who paid for it? NCCIH suggests asking whether the sponsor or investigators have a financial or reputational "stake" in the outcome. [12]
• Negative results count. A result of "no benefit" is evidence too, as with ginkgo for dementia. [2]

Our guide to reading a herb study or news headline goes through these steps with examples, and how herb trials are designed explains why results differ.

What is not known

Proper evidence not found for a single measure of "how many herbs work". We looked at the NCCIH herb pages and the Cochrane review cited here, and each answers one herb at a time. This page therefore makes no overall claim.

Traditional use is a different kind of evidence. It tells you how people have used a plant, not whether a trial confirmed it. See traditional use versus research evidence.