Forty-seven patients with unilateral hip disease had a preoperative and postoperative gait analysis of total hip arthroplasty with a follow-up period of five to ten years. The object was to correlate the levels of function with mechanical loosening. Unilateral total hip arthroplasty in uncomplicated patients restored function to 80% of normal over a two- to four-year period. Patients followed longer than five years appeared to have an accelerated decline in function unassociated with pain. No Class C patient, regardless of age or gait velocity, developed loosening or required revision of the hip prosthesis. Survivorship studies revealed an 85%-90% incidence of loosening by the seventh postoperative year in Class A patients 55 years or younger and those walking preoperatively at more than 50% of normal gait velocity. In these same two groups of patients, the probability of revision for mechanical failure by ten years is 58% for age and 66% for fast walkers. The majority of patients in this study had a technical rating not acceptable by present standards so that these results represent most likely the worst prognosis for longevity of total hip arthroplasty. In high-risk patients, protected ambulation with crutches for six months postoperation may improve their prognosis.