Study Design Historical cohort, including selected subgroups. Objectives To understand the long-term effects of exercise on back-related outcomes, back pain, sciatica, back-related hospitalizations, pensions, and magnetic resonance imaging findings were studied among former elite athletes. Summary of Background Data Exercise and sports participation have become increasingly popular, as have recommendations of exercises for back problems, but little is known about their long-term effects. Methods Questionnaires were returned by 937 former elite athletes and 620 control subjects (83% response rate). Identification codes allowed record linkage to hospital discharge and pension registers. Magnetic resonance images were obtained of selected subgroups with contrasting physical loading patterns. Results Odds ratios for back pain were lower among athletes than among control subjects, with significant differences in endurance, sprinting and game sports, and wrestling and boxing. No differences in the occurrence of sciatica or in back-related pensions and hospitalizations were seen. When comparing lumbar magnetic resonance images of 24 runners, 26 soccer players, 19 weight lifters, and 25 shooters, disc degeneration and bulging were most common among weight lifters; soccer players had similar changes in the L4-S1 discs. No significant differences were seen in the magnetic resonance images of runners and shooters. Conclusions Maximal weight lifting was associated with greater degeneration throughout the entire lumbar spine, and soccer with degeneration in the lower lumbar region. No signs of accelerated disc degeneration were found in competitive runners. However, back pain was less common among athletes than control subjects and there were no significant differences in hospitalizations or pensions. No benefits were shown for vigorous exercise compared with lighter exercise with respect to back findings.