The purpose of the study was to assess the effect of cigarette smoking on indicators of bone strength across a wide age range, controlling for physical activity and neuromuscular performance. We conducted a cross-sectional study with 41 smokers (mean age +/- SD, 41.0 +/- 16.1 yr) and 53 nonsmokers (47.5 +/- 18.2 yr) of both sexes. Bone strength indicators (BSI) were assessed in the lower leg and forearm by peripheral quantitative computed tomography along with physical activity, muscle cross-sectional area, and maximal voluntary muscle force. Physical activity level and muscle cross-sectional area of the leg and arm were similar in smokers and nonsmokers. Although trabecular volumetric bone mineral density and epiphyseal bone mineral content, both indicators of BSI, decreased with age in the nonsmokers' tibia (P < 0.001), this was not observed in the smokers (interaction age x smoking: P = 0.014 and P = 0.032 for density and content, respectively). Regression coefficients were nonsignificant in nonsmokers, whereas coefficients in smokers were -1.24 mg/cm x yr [95% confidence interval (CI) = -2.16-0.33; P = 0.01] for content and -1.20 mg/cm(3) x yr (95% CI = -1.76-0.62; P < 0.001) for trabecular density. The BSI values in the smokers were independent of their smoking history (r(2) = 0.000-0.021), and no effects of sex were observed in the smoking-related differences in BSI. Smoking compromises bone strength by diaphyseal marrow cavity expansion and epiphyseal trabecular bone content reductions. These effects seem to wane with age. The causes of the attenuated effect of smoking on bone at old age remain enigmatic but might be linked to an interaction between the smoke-related factors and senescence processes affecting bone.