The objective of this study was to compare comorbidity, functional ability, and health care utilization in veterans with total knee arthroplasty (TKA) or total hip arthroplasty (THA) versus matched control populations. A cohort of veterans using Veterans Affairs (VA) healthcare system reported limitations in six activities of daily living (ADLs; bathing, dressing, eating, walking, transferring, and using the toilet), demographics, and physician-diagnosed comorbidity. VA databases provided healthcare utilization and International Classification of Diseases-9/Common procedure terminology codes for TKA/THA. Patients were classified as: (1) primary TKA; (2) primary THA; (3) combination group (>or=1 procedure); and (4) control veteran population (no THA/TKA). Multivariable regression analyses compared the risk or counts of ADL limitation and in-/out-patient visits. After multivariable adjustment, TKA, THA or combination groups had significantly higher prevalence of the following compared to veteran controls: arthritis, diabetes, or heart disease (p < 0.0001 each), severe (>or=3) ADL limitation (33%, 42%, 42% vs. 24%; p < 0.0001), and annual hospitalization rate (24%, 19%, 26% vs. 16%, p < 0.0001). Annual outpatient surgery visits were more (2.5, 2.3, 2.3 vs. 2, p = 0.01) and risk of any mental health outpatient visit was lower (12%, 11%, 12% vs. 18%, p = 0.0039). All ADLs, except eating, were significantly more limited in arthroplasty groups (p or=65 years; and 1.2-, 1.6-, and 4-fold in >or=85, 75-84, and 65-74 years. Poorer function and higher comorbidity and utilization in veterans with TKA/THA suggest that this group is appropriate for interventions targeted at improving function and decreasing utilization.