Questionnaires are commonly used in orthopaedic outcome studies. This study sought to determine if responses to a simple standardized questionnaire correlated with responses obtained during a physician interview in evaluation of clinical outcome following hip and knee arthroplasty. One hundred sixty-two patients with 201 hip and knee arthroplasties were asked to fill out a questionnaire prior to returning for routine follow-up evaluation. There was a highly significant correlation (P < .0001, r = .74) between scores calculated from patient responses on the questionnaire and those calculated from responses recorded during the subsequent physician visit. There was no significant difference between patient and physician clinical hip scores, but physicians gave significantly higher knee scores than patients for both long- ( > 4.5 years, P < .05) and short-term ( < or = 4.5 years, P < .0001) follow-up periods; however, 97% of patient responses were within one grade of physician-recorded answers to the same questions. Eight and one-half percent of scores differed in overall evaluation from good-excellent to fair-poor categories. This study both validates and defines more clearly the limitations of questionnaires for follow-up evaluation of clinical results following total hip and knee arthroplasty.