Little is known about the influence of long-term daily physical activity (PA) on lumbar bone mass after peak bone mass has been reached, that is, during [young] adulthood. The purpose of this study was to investigate the longitudinal relationship between PA and lumbar bone mineral density (LBMD) in healthy subjects over a 10-year period. The data reported here relate to 225 male and 241 female participants in the Amsterdam Growth and Health Longitudinal Study, who were measured at the mean ages of 27, 32, and/or 36. LBMD, habitual daily PA, total body weight, and calcium intake were assessed at each measurement point. The effects of two aspects of PA were analyzed: the mechanical (MECHPA; sum of all ground reaction forces) and metabolic (METPA; weighted metabolic score of intensity, frequency, and duration) components, each within a separate model. Multilevel analysis was used to investigate the relationship between PA and LBMD over the 10-year period. Gender, total body weight, and calcium intake were included in the analysis as covariates. The results indicated that MECHPA was a significant positive linear predictor of LBMD for males (r = 0.09; p < 0.001) but not for females. For the METPA, no linear longitudinal relationship with LBMD was found. The results suggest that there is a metabolic threshold at which extra PA becomes "deleterious" and METPA in its totality becomes ineffective for LBMD. It is concluded that during the (young) adult period, between 27 and 36 years of age, PA causing mechanical loading on the skeleton has a small positive influence on LBMD in males.