Using the population based data resources of the Rochester Epidemiology Project and a postal survey, we estimated all costs (direct medical, indirect medical, and nonmedical) incurred among a population based prevalence cohort of individuals with osteoarthritis (OA) and a nonarthritic comparison cohort. Cost estimates were compared between the 2 groups adjusting for age and sex differences. Our results demonstrated that people with OA incur substantial incremental medical costs, not only for arthritis care, but also for the care of neurological, gastrointestinal, cardiac, and respiratory conditions compared to nonarthritics. These data also demonstrated important and statistically significantly higher levels of work disability among people with OA compared to their age and sex adjusted peers without arthritis from the same community.