The role of drinking water fluoride content for prevention of osteoporosis remains controversial. Therefore, we analyzed the influence of drinking water fluoridation on the incidence of osteoporotic hip fractures and bone mineral density (BMD) in two different communities in eastern Germany: in Chemnitz, drinking water was fluoridated (1 mg/L) over a period of 30 years; in Halle, the water was not fluoridated. BMD was measured in healthy hospital employees aged 20-60 years (Halle: 214 women, 98 men; Chemnitz: 201 women, 43 men, respectively) using dual-energy X-ray absorptiometry. Hip fractures in patients > or = 35 years admitted to the local hospitals in the years 1987-1989 were collected from the clinic registers. There was no difference in age, anthropometric, hormonal, or lifestyle variables between the two groups. Mean fluoride exposure in Chemnitz was 25.2 +/- 7.3 years. No correlation was found between fluoride exposure and age-adjusted BMD. We found no significant difference in spinal or femoral BMD between subjects living in Halle and Chemnitz [lumbar spine: 0.997 +/- 0.129 (g/cm2) vs. 1.045 + 0.171 (g/cm2), p = 0.08, for men; 1.055 +/- 0.112 (g/cm2) vs. 1.046 +/- 0.117 (g/cm2), p = 0.47, for women]. The fracture incidence showed an exponential increase with aging in men and women with an incidence about 3.5 times higher for women. In Chemnitz, we calculated an age-adjusted annual incidence of 142.2 per 100,000 for women and 72.5 per 100,000 for men, respectively. In Halle, the incidences were 178.5 per 100,000 for women and 89.2 per 100,000 for men. There was a lower hip fracture incidence after the age of 85 in women in Chemnitz (1391 per 100,000 in Chemnitz vs. 1957 per 100,000) in Halle, p = 0.006). Using the age-adjusted incidences, significantly fewer hip fractures occurred in Chemnitz in both men and women. In conclusion, our study suggests that optimal drinking water fluoridation (1 mg/L), which is advocated for prevention of dental caries, does not influence peak bone density but may reduce the incidence of osteoporotic hip fractures in the very old.