Long-term effects of high dietary protein intake on muscle and bone structure in the elderly are not clear. The aim of this study was to investigate the relationship between baseline protein intake and lean mass and BMC 5 yr later in a cohort of elderly postmenopausal women. A total of 862 community-dwelling women 75 +/- 3 yr of age provided baseline data including nutrient intake assessed by a food frequency questionnaire. At 5 yr, upper arm muscle area (UAMA) and body composition using DXA were measured. Baseline protein intake was 81 +/- 28 g/d (1.2 +/- 0.4 g/kg/d), contributing 19 +/- 3% of total energy intake. There were positive correlations between baseline protein intake and whole body and appendicular bone-free lean mass and BMC (r = 0.14-0.18, p < 0.001) and UAMA (r = 0.08, p < 0.05). Compared with those in the lowest tertile of protein intake (<66 g/d), women in the top tertile (>87 g/d) had 5.4-6.0% higher whole body and appendicular lean mass and UAMA and 5.3-6.0% higher whole body and appendicular BMC. These effects remained after adjusting for potential confounders. However, the effect on BMC disappeared after further adjustment for lean mass. This study shows that high protein intake is associated with long-term beneficial effects on muscle mass and size and bone mass in elderly women. The protein effect on bone may be partly mediated by its effects on muscle.