Accepted outcome measures in randomized controlled trials (RCTs) in osteoarthritis (OA) include patient-reported assessments of physical function and health-related quality of life (HRQOL). Available data can inform treatment decisions when statistically significant changes are viewed in terms of clinically important improvements. Patient-reported outcomes validated in OA include global assessments of pain, disease activity, and disease-specific and generic measures of physical function and HRQOL. Definitions of minimum clinically important differences (MCID) have been derived from RCTs with physical therapy, nonsteroidal anti-inflammatory drugs (NSAIDs), and cyclooxygenase-2 selective agents. Definitions of MCID should serve only as guidelines based on mean changes in a treatment group, and do not necessarily reflect clinically meaningful improvements for an individual patient. They help to interpret data across treatments and patient populations. Definitions of MCID may differ for the type of intervention assessed; additional methodologic issues must be addressed when evaluating nonpharmacologic treatments. Based on RCTs in OA evaluating physical therapy, cyclooxygenase-2 agents, and NSAIDs, the Western Ontario and McMaster Osteoarthritis Index is valid, reliable, sensitive to change, and correlates closely with the generic Medical Outcomes Survey Short-Form 36 measure of HRQOL. When evaluating RCT data, understanding derivation and MCID values of outcome measures facilitates informed therapeutic decisions regarding therapeutic interventions.