This was a retrospective study reviewing 68 hip arthroplasties performed with noncemented, porous-coated components at three institutions in the period between December 1983 and June 1987. Fifty of these were available for follow-up study. The mean follow-up period was three years (range, 18-56 months). Thirty-seven patients were treated by total hip arthroplasty (THA) with porous-coated components, and thirteen were implanted with porous-coated femoral stems with bipolar endoprostheses using a porous-coated femoral stem and a universal head. The mean Harris hip score at follow-up evaluation for the entire group was 85.7 points. Eighty percent of the patients had either a good or excellent result. Those patients implanted with a porous femoral stem with bipolar acetabular components had only 54% good or excellent results compared with 90% good or excellent results for those having THA. Roentgenographic findings included a 59% rate of femoral neck cortex osteopenia, an 80% rate of medial or lateral distal cortical hypertrophy, and an 80% rate of distal bone bridging. All of these roentgenographic changes were secondary to changes in stress transference to the proximal femur. There was no correlation between thigh pain or the degree of bone-prosthesis radiolucencies and the fit of the femoral stem.