How Herb Clinical Trials Are Designed and Why Results Differ

A good herb trial compares a defined herbal product with an inactive look-alike in people who do not know which one they are getting. Herbs are harder to test than drugs made of one compound, because the product itself can vary. That is one reason trials of the same herb often disagree. This page explains the design and the herb-specific problems. It does not recommend any herb.

The basic design

• Control group. NCCIH describes a controlled trial as one that compares groups of participants who are as identical as possible in other respects. One group receives the treatment and the other receives standard care or an inactive substance. [1]
• Blinding. In a double-blind study neither the participants nor the researchers know who is in which group. [1] NCCIH says it can be surprisingly difficult to avoid bias in clinical trials. [2]
• Size. Small sample sizes are more likely to produce results due to chance. [3]
• Meaning of the result. Differences can be statistically significant but so small that they are not clinically significant for patients. [4]
• Where trials sit. NCCIH says clinical trials give the clearest information about whether a treatment is effective and safe in humans. [5] Studies in cells or animals are preclinical and cannot show this.

Where herb trials get harder

The product has to be described. A review of 206 randomized trials of ten herbal medicines found that 45% of the reporting checklist items were reported across all trials, and that key herbal details such as the percentage of active constituents and the type or form of the preparation may be missing. [6] The reporting guideline CONSORT applies to herbal products, [7] and a version specific to herbal medicine trials was written for authors. [6]

The placebo is hard to make. A 2026 survey of 136 trials of Chinese herbal medicine found that deficiencies persist in how placebos were designed and reported. In 33 of the trials the placebos contained traditional Chinese medicine ingredients, which the authors say challenges the principle that a placebo is pharmacologically inert. [8] (The paper has been corrected by an erratum, so check the journal version for the figures.)

Many different products carry one name. The Cochrane review of echinacea for colds found 24 double-blind trials that used different species and parts of the plant. Because of strong clinical heterogeneity (the studies were too different), the authors did not pool them for the main analysis. [9]

The product in the trial may not be the product on the shelf. NCCIH says supplements you buy may differ in important ways from products tested in studies. [10]

What the trials found, as examples

Limits

Proper evidence not found for a single rule that explains all herb trial differences. We searched the NCCIH, Cochrane and PubMed sources cited here. Product variation, placebo design, trial size and funding each play a part and none alone explains the pattern.