Sioux Indians swore by it for snakebite, the Comanche used it for toothache, while early settlers adopted it as a home remedy for colds. Now echinacea, or purple coneflower, seems to be making a comeback, particularly in the US where annual sales are nearly $80m (pounds 50m). Enthusiasts claim the plant - extensively hyped in the British press this winter - has anti-viral properties, is particularly effective at warding off this year's flu strain, and can also speed the rate of recovery.

In addition, it is claimed that echinacea has antifungal and antibacterial ingredients: one of them, echinacein, is responsible for the plant's highly pungent taste and is thought to soothe a sore throat. The herb is also being tried for more serious conditions involving immune failure, including chronic fatigue syndrome and, more controversially, Aids.

So what is the evidence? Medical herbalists say that the plant (which comes in tablet, root or tincture form and is available over the counter, often in combination with other herbs) has been shown by laboratory research to have a wide set of actions on the immune system. They point out that it contains complex sugars called polysaccharides, and chemicals called polyacetylenes, both of which stimulate the production of interleukin and interferons, proteins crucial to the immune response. They claim that the plant is particularly helpful at combating flu as the propensity of the virus to mutate makes vaccines unreliable.

All of which sounds impressive enough: undoubtedly, some plants are brilliant chemical factories, which is why so many of our most potent medicines are derived from them. But in the laboratory, thousands of different molecules have been found to interfere with viruses, and so far there is little evidence about which, if any, echinacea might contain.

Herbalists would argue that many ingredients contribute to the plant's total effect and that not knowing the active ingredient does not matter - as long as the remedy works. But does it? In Germany, where echinacea has been popular for years, researchers think that its effect has been overestimated (ironic since the Germans are always being quoted by British herbalists in search of credibility). A major review of 26 clinical trials in which the herb was used to treat infections found most to be of poor quality and the results inconclusive. Two more recent studies of its effect on colds and flu have not yet been published, but, according to the author, the results are not very convincing. And although randomised trials of the herb on healthy volunteers, in which antibody levels in the blood were measured, suggests that it might improve immune function, again, the results were not consistent.

With hard evidence so shaky, there is little point in taking echinacea either to avoid getting colds or flu or to treat them. If, on the other hand, you catch several colds a year and take a long time to recover, your immune system is probably not as robust as it might be. In which case echinacea might help - but then so might more sleep, less stress and a better diet.

Those in the know about flu, incidentally, are putting their money on a new anti-flu drug being developed by Glaxo. Based on the molecule GG167, the drug has been shown to be clinically active in controlled trials and could be on the market in a couple of years.