Genetic factors play an important role in determining bone mass and several genes are involved in this process. Interleukin-6 (IL-6) is a candidate gene for regulation of bone mineral density (BMD) and it has been suggested recently that novel IL-6 -174 G/C allelic variants may be associated with peak BMD in young men and with bone resorption in elderly women. In this study, we assessed the relationships between IL-6 gene polymorphism, peak BMD, rate of postmenopausal BMD loss, and bone turnover in women. BMD was measured by dual-energy X-ray absorptiometry in 255 healthy premenopausal women, aged 31-57 years. BMD loss at the forearm was measured over 4 years in 298 healthy untreated postmenopausal women, 50-88 years (mean 64 years). We also measured levels of serum osteocalcin, bone alkaline phosphatase, and N-propeptide of type I collagen for bone formation and three markers of bone resorption, including urinary and serum C-terminal cross-linking telopeptide of type I collagen and urinary N-terminal telopeptide of type I collagen, in both pre- and postmenopausal women at baseline. In premenopausal women we found a significant association between IL-6 genotypes and BMD at the whole body (analysis of variance [ANOVA], p = 0.03), femoral neck (p = 0.03), trochanter (p = 0.014), Ward's triangle (p = 0.03), and total hip (p = 0.006), with subjects having the CC genotype showing 3%-7% higher BMD levels than their GG counterparts. However, after matching women with CC and GG genotypes for body height the differences decreased (2%-4%), and were no longer significant (p = 0.10-0.23). In postmenopausal women the mean rate of loss at the ultradistal radius was significantly associated with IL-6 genotypes (ANOVA, p = 0.049), with women having the CC genotype showing a significantly greater rate of bone loss (p < 0.05) compared with their GC and GG counterparts. After adjustment for weight changes, the difference in the rate of ultradistal radius bone loss between genotypes decreased and was not significant (p = 0.06 for CC vs. GG). A similar trend was observed for distal radius bone loss (p = 0.10, ANOVA), but not for the middle radius. We found no significant association between genotypes, bone turnover markers in premenopausal women, and either bone turnover or BMD in postmenopausal women. We conclude that this new functional IL-6 polymorphism was weakly associated with level of peak BMD and the rate of forearm trabecular postmenopausal bone loss in this cohort of healthy French women. IL-6 genotypes accounted only for a small proportion of the interindividual variation of both peak BMD and rate of bone loss and were not significant after adjustment for height and changes in body weight, respectively, suggesting that part of the effect may have been due to the differences in body size. Larger long-term studies are necessary to assess adequately the relationships between IL-6 genotype, rate of bone loss, and risk of fracture.