You may recall Dr Barrett's article on the "science" behind homeopathy, exposing it in all its ludicrousness. After reading it, I suppose a homeopath could still mutter something along the lines of, "Well, just because science has yet to discover the principle behind homeopathy doesn't meant it doesn't work."
Of course, I would argue that it isn't so much that science is yet to discover some new set of forces than it is that science is saying that homeopathy is rubbish. However, even if we grant them this point (a massive concession), we can still know that homeopathy isn't any better than a placebo. How? We do trials and see.
Of course, the trials must be of very high quality, or else the results we get from them won't be trustworthy. Homeopaths love to quote some trial from the 1990s of abysmal quality in order to make themselves sound scientific (usually while simultaneously degrading the scientific method; go figure).
But if we adhere to strict quality control guidelines in the trials we conduct, the result is repeatedly the same: homeopathy is no better than a placebo.
Note, incidentally that being no better than a placebo doesn't mean that those who take homeopathic remedies don't sometimes get better. It just means that it's no better than sugar water fed to a credulous child. And so it probably shouldn't be taken seriously when it comes to tax money distribution, etc.
All these points are outstandingly dealt with in this blog entry by Ben Goldacre, a pugnacious defender of rationality (especially within the health field) from his dual vantage points of the columns of the (British) Guardian newspaper and his blog, Bad Medicine. Enjoy - the other articles are also very much worth a read.