Aims: There is a need to better understand work incapacity due to musculoskeletal disorders (MSD) and the factors that contribute to being granted disability pension (DP) with such disorders. A twin cohort study would serve a powerful tool responding to this knowledge gap by providing information on factors affecting DP when controlling for family background. The purpose was to investigate the incidence of and risk factors for DP due to any MSD ( n = 1,819) and specifically due to osteoarthritis (OA, n = 677) in a twin cohort of 24,043 people over a 30-year follow-up. Methods: Data on twin pairs from a mailed questionnaire during the baseline year of 1975 were followed up with register data regarding DP, emigration, old-age pension, and death. For statistical analysis, univariate and multivariate Cox proportional hazard ratios were estimated. Results: Baseline musculoskeletal pain, frequency of use of analgesics, body mass index, and chronic diseases, as well as education and social class were significant risk factors for DP due to MSD at follow-up in both sexes. These factors were also the significant predictors of DP due to OA in men. In women, DP due to OA was best predicted by baseline musculoskeletal pain and lower social class. Conclusions: The risk of DP due to MSD and OA seemed to be influenced by comorbidities, educational level and social class. Analyses of twin pairs discordant for DP confirmed the results. Accumulated health problems and chronic conditions during the life course may lead to permanent work incapacity.