Targeted weight-bearing activities during the pre-pubertal years can improve cortical bone mass, structure and distribution, but less is known about the influence of habitual physical activity (PA) and fitness. This study examined the effects of contrasting habitual PA and fitness levels on cortical bone density, geometry and mass distribution in pre-pubertal children. Boys (n = 241) and girls (n = 245) aged 7–9 years had a pQCT scan to measure tibial mid-shaft total, cortical and medullary area, cortical thickness, density, polar strength strain index (SSIpolar) and the mass/density distribution through the bone cortex (radial distribution divided into endo-, mid- and pericortical regions) and around the centre of mass (polar distribution). Four contrasting PA and fitness groups (inactive–unfit, inactive–fit, active–unfit, active–fit) were generated based on daily step counts (pedometer, 7-days) and fitness levels (20-m shuttle test and vertical jump) for boys and girls separately. Active-fit boys had 7.3–7.7 % greater cortical area and thickness compared to inactive–unfit boys (P <0.05), which was largely due to a 6.4–7.8 % (P <0.05) greater cortical mass in the posterior–lateral, medial and posterior–medial 66 % tibial regions. Cortical area was not significantly different across PA-fitness categories in girls, but active-fit girls had 6.1 % (P <0.05) greater SSIpolar compared to inactive–fit girls, which was likely due to their 6.7 % (P <0.05) greater total bone area. There was also a small region-specific cortical mass benefit in the posterior–medial 66 % tibia cortex in active-fit girls. Higher levels of habitual PA-fitness were associated with small regional-specific gains in 66 % tibial cortical bone mass in pre-pubertal children, particularly boys.