Fluoride has been proven to prevent caries, as such dental health has improved dramatically since the addition of fluoride compounts to drinking water. HOWEVER, too much can be bad.
There was a good article in a recent Scientific American which outlined the findings of recent research into the physiological effects of fluoride. Specifically: there is a risk that excessive fluoride consumption can increase the risks of disorders affecting teeth (especially in young people, fluoride can disrupt the formation of permanent teeth, resulting in discolouration and disfiguration), bones (causes the proliferation of osteoblasts, increasing the size of, but decreasing the strength of, bones. The result: increased incidence of osteoperosis. They MIGHT cause cancers too), the brain (fluoride can accumulate in the brain and cause hyperactivity in young rats) and the thyroid. All of these studies use fluoride in concentrations much higher than those found in drinking water - that's not to say that low concentrations don't have any effect, but it would be foolish to draw parallels between the two.
The only conclusion that can be drawn with current evidence, however, is that some fluoride is good to prevent caries and increase dental health, whereas too much is probably bad. Decreasing the average fluoride concentration in US drinking water from 0.71ppm to 0.2-0.3ppm would probably suffice. In addition, if the government got off their arses and pumped more money into research concerning the physiological effects of fluoride, then we can decrease the negative effects while still preventing caries until we have more evidence.
Concentration of fluoride in common foods/drinks (parts per million):
Brewed black tea: 3.73
Raisins: 2.34
White wine: 2.02
Apple juice: 1.09
Brewed coffee: 0.91
US drinking water (average): 0.71