Our intent was to evaluate clinically total hip arthroplasty in the dog using the same procedures and prostheses. We replaced 20 hips in 15 dogs of both sexes and various breeds, weights, and ages. All animals exhibited pain caused by degenerative joint disease of one or both coxofemoral joints. Using Amstutz's system of grading at a minimum of 12 months post-operatively, none of the hips replaced were graded poor, 12.5% were graded fair, 12.5% good, and 75% excellent. Loosening of prosthetic components was by far the most common problem encountered in this study. Luxations were virtually eliminated by the use of the Richards Canine II total hip replacement system. Infections were found to be directly proportional to surgery time, number of operations performed, or both. Component failure occurred only when the size of the prosthesis was not properly matched to the size of the dog. Although these problems still must be overcome, we conclude that total hip arthroplasty is a valuable adjunct to veterinary medicine in selected cases.