Between March 1984 and March 1989, thirty-four patients who had an infection at the site of a cemented total hip prosthesis were managed with resection arthroplasty and delayed implantation of a porous total hip prosthesis without cement. The interval from the time of the resection arthroplasty to the implantation of another prosthesis averaged eight months (range, three to nineteen months). At an average of forty-seven months (range, twenty-four to seventy-two months) after the reimplantation, six patients (18 per cent) had recurrence of the infection. Patients who had rheumatoid arthritis were at significantly higher risk for the development of a recurrent infection (p < 0.01). Of the twenty-eight patients who did not have a recurrent infection, six had definite radiographic evidence of loosening of the femoral component at the latest follow-up evaluation. For twenty-five of the twenty-eight patients, sufficient data were available for calculation of the Mayo Clinic hip score; only fourteen (56 per cent) of these patients had a satisfactory functional outcome. The high (68 per cent) rate of complications and the long-term durability of the prosthesis in these patients remain a concern. The fact that 18 per cent of the patients had a recurrent infection suggests that avoidance of the use of bone cement does not improve the rate of resolution of infection after a delayed revision operation in patients who have an infection following a total hip arthroplasty.