This prospective observational study investigated the relationship between the length of hospital stay (LOS) and outcomes at 3 months for primary total hip arthroplasty for osteoarthritis. Mean length of postoperative stay was 9.5 +/- 2.8 days. Predictors of LOS were patient's age, sex, and number of comorbidities; preoperative Charnley scores and Nottingham Health Profile measures; complications; and hospital in which surgery took place. LOS was found to have a small negative correlation with outcome. The dominant association with improved outcome was the severity of the patients' impairment preoperatively. These data suggest that in situations in which adequate rehabilitation and support are available after discharge, a marginal reduction in postoperative LOS--from the average of 10.3 days observed at 1 hospital to the average of 8 days observed at another--would not adversely affect the short-term outcome.