OBJECTIVE To examine demographic and disease-related variables that affect pain in a large cohort of patients with juvenile rheumatoid arthritis (JRA). METHODS Selection criteria were an onset of JRA >/=5 years prior to study and age >/=8 years at the time of the study. Pain was measured by a self-administered 10-cm visual analog scale. Possible explanatory variables studied included age at study, sex, race, onset subtype, active disease duration, active joint count, and physician's global assessment (PGA). RESULTS In a multiple regression model, active disease duration, PGA, and age at study were independent predictors explaining 22% of the variation in pain scores. Stratified analyses showed an effect of age in the 8-15-year group, but not in older patients. CONCLUSIONS Disease-related factors explain only a small proportion of the variation in pain scores. Age has an effect on pain scores only in younger patients. The role of other factors, including psychosocial factors, needs further study.