OBJECTIVE: To investigate recovery from total hip replacement over a three-year period on the basis of patient perceptions of health-related quality of life, demographic (age, gender, family support) and clinical characteristics (co-morbidities, hospital admissions), use of and satisfaction with health services, unmet health needs and social re-engagement. DESIGN: Telephone survey of patients' scores on the the 36 item Short Form health survey (SF 36) at three-years as a basis for comparison with scores 12 weeks after discharge; analysis of survey responses on demographic and clinical variables, health services use and satisfaction, unmet health needs and resumption of pre-operative social activities. SETTING: One tertiary hospital in South East Queensland, Australia. PARTICIPANTS: Sixty-two total hip replacement patients from an original cohort of 95 participants in a study three years previously to investigate these variables at one, two, four, eight and 12 weeks post-discharge. MAIN OUTCOME MEASURES: SF36 scores, survey responses on number and type of co-morbidities, age, family support, type and frequency of health services used, satisfaction with services, hospital admissions, resumption of social activities. RESULTS: No significant differences were found on any SF-36 scores, but General Health had declined. Women's Physical Functioning scores fell below population norms, men's remained above the norms. Physical composite scores showed a decline with age, and the Mental Composite Scores increased with age. The number of co-morbidities had increased over three years, with 58% being admitted to hospital. Half reported unmet health needs, related primarily to non-hip problems. Only general practitioner services were used monthly or more, with satisfaction ratings remaining high for all services used. Participation in social activities was increased from prior to surgery for 42% of participants. CONCLUSIONS: Recovery after hip replacement surgery is dramatic, especially in alleviation of pain, but for older patients, there is a subsequent decline in general health concomittant with others in this demographic group. Differences in men's and women's patterns of recovery suggests differential planning to provide more realistic expectations for recovery and aftercare.