- While bone mass and geometry are largely genetically determined, mechanical loading is considered to be an important additional determinant. This study investigates to what extent very high mechanical loading begun at a young age and sustained afterward can affect tibia bone mass and geometry in middle age. Cohorts from a common ethnic background, with a history of very high and very low tibia bone loading based on an assessment of their activities according their strain levels were compared. The study hypothesis was that the tibia bone density and geometric strength parameters would be greater in the high bone loading cohort. Subjects from a group of elite infantry recruits who sustained a 31% incidence of stress fractures during their basic training in 1983, were reviewed 25 years later. The tibia bone strength of 25 of these soldiers, 11 of whom had sustained stress fractures, was compared to a group of 20 subjects who received exemption from military service in 1982–5 because they were religious scholars and who continued these studies afterwards. Anthropometric measurements were made. The bone density and geometric strength of the tibia was assessed by quantitative computerized tomography (QCT). The average daily dietary intake and metabolic expenditure of subjects were assessed by questionnaires. At the 25 year follow-up soldiers were on an average 3 cm taller than the religious scholars (p = 0.02) and had lower abdominal girths (p = 0.03). There was no difference in the tibia cortical density between cohorts in spite of the fact that the religious scholars had lower daily calcium intakes (p = 0.02). Soldiers had stronger tibias based on geometric engineering criteria. The mean area moments of inertia (p = 0.02, p = 0.04) and polar moments of inertia (p = 0.02) were 16% larger in the soldier cohort. By multivariate regression analysis greater height, weight and daily energy expenditure were related to larger bone geometric strength parameters. According to semipartial eta-square analysis, between 39% to 45% of the variance in the area moments of inertia between the cohorts was attributable to these three parameters. The religious scholars burned less calories daily, principally because they did no sport activity (p = 0.001). There was no difference in tibia bone strength parameters between soldiers who did and did not sustain stress fractures in their 1983 basic training. In conclusion, in a middle age population with a common ethnic origin, the high bone loading cohort had stronger tibias than the low bone loading cohort based on larger geometric strength properties and not because of higher cortical density. In spite of being at the extremes of the bone loading spectra, the tibia area moment of inertia of the two cohorts in this study differed by only 16%, with part of this difference attributable to factors other than bone loading. We do not know for sure if the difference in the geometric properties is related to high bone loading or whether people with stronger bones are more likely to engage in high bone loading. Healthy male subjects who sustained stress fractures at a young age do not have weaker tibias at middle age according to QCT measurements.