In order to investigate the extent to which musculoskeletal sickness absence was influenced by a range of circumstances concerning family background and health in early life, we established a register-based cohort of all live-born in Norway between 1967 and 1976. Personal data on parental factors and health early in life were recorded prospectively from birth onward in the Medical Birth Registry of Norway, the National Insurance Administration, Statistics Norway, and the Central Population Register. We collected data in the National Insurance Administration on the first spell of medically certified long-term (>16 days) musculoskeletal (International Classification of Primary Care group L) sickness absence in 2000-2003 among 378, 356 participants who were considered to be at risk of sickness absence on January 1st, 2000. The 4-year musculoskeletal absence risk was 0.264 for women and 0.156 for men. Parental education level was associated with musculoskeletal sickness absence, with increasing adjusted relative risks by decreasing educational level for both genders. Associations with other early determinants (birth weight, childhood disease, parental survival, parental disability, parental income, and parental marital status) were all close to unity. Parental education level attributed 36% (95% confidence interval 33-38) to the population risk for women and 67% (64-70) for men. The parental education association was partly mediated through own educational attainment, which was strongly associated with musculoskeletal sickness absence in itself. Our data suggest that mechanisms acting early in life could influence later risk of musculoskeletal sickness absence.