Musculoskeletal pain that affects multiple body sites is typically regarded as comorbidity to single-site pain. Pain present in multiple sites, however, is more severe and disabling compared to single-site pain. This study aimed to prospectively investigate the change in the number of pain sites over 14years, in addition to identifying predictors of multi-site pain. In 1990 and 2004, questionnaires about musculoskeletal pain were mailed to six birth cohorts in Ullensaker, Norway. Data on demographic, lifestyle, and health-related variables were also collected. Participation rate in 1990 was 67.2% and among those 60.4% participated in 2004. A slight increase in the average number of pain sites occurred between 1990 and 2004, but results showed a relatively stable pattern of pain reporting over a period of 14 years. Several demographic, lifestyle, and health-related variables in 1990 predicted the number of pain sites at follow-up in the bivariate analyses. However, only sex, age, sleep quality, and educational level remained significant in the final multivariate model after controlling for the number of pain sites at baseline. The final model explained 35% of the variance, of which nearly 80% was accounted for by the number of pain sites at baseline. As the pattern of reporting the number of pain sites appears relatively stable across adulthood and baseline multi-site pain demonstrated strong predictive utility, studies investigating the occurrence of multi-site pain in children and adolescents are recommended to determine potential causal factors contributing to the early course and development of multi-site musculoskeletal pain.