Background: Physical activity (PA) increases bone mass, especially in late prepuberty and early puberty, but it remains unclear if and how PA affects both bone formation and bone resorption. Materials and Methods: We included 191 boys and 158 girls aged 7.7 ± 0.6 (mean ± SD) in a population-based PA intervention study. The intervention group (123 boys and 94 girls) received daily physical education (PE) in school (40 min/day; 200 min/week) from study start and during the nine compulsory school years in Sweden. The controls (68 boys and 64 girls) received the national standard of 1–2 classes PE/week (60 min/week). During the intervention, blood samples were collected at ages 9.9 ± 0.6 (n = 172; all in Tanner stages 1–2) and 14.8 ± 0.8 (n = 146; all in Tanner stages 3–5) and after termination of the intervention at age 18.8 ± 0.3 (n = 93; all in Tanner stage 5) and 23.5 ± 0.7 (n = 152). In serum, we analyzed bone formation markers [bone-specific alkaline phosphatase (bALP), osteocalcin (OC), and N-terminal propeptide of collagen type 1 (PINP)] and bone resorption markers [C-terminal telopeptide cross links (CTX) and tartrate-resistant acid phosphatase (TRAcP 5b)]. Linear regression was used to compare age and sex-adjusted mean differences between intervention children and controls in these markers. Results: Two years after the intervention was initiated (at Tanner stages 1–2), we found higher serum levels of bALP and OC, and lower serum levels of TRAcP 5b in the intervention compared with the control group. The mean difference (95% CI) was for bALP: 13.7 (2.1, 25.3) μg/L, OC: 9.1 (0.1, 18.1) μg/L, and TRAcP 5b: −2.3 (−3.9, −0.7) U/L. At Tanner stages 3–5 and after the intervention was terminated, bone turnover markers were similar in the intervention and the control children. Conclusion: Daily school PA in the late prepubertal and early pubertal periods is associated with higher bone formation and lower bone resorption than school PA 1–2 times/week. In late pubertal and postpubertal periods, bone formation and resorption were similar. Termination of the intervention is not associated with adverse bone turnover, indicating that PA-induced bone mass benefits gained during growth may remain in adulthood.