The results of revision elbow arthroplasty with use of the semiconstrained Mayo-modified Coonrad implant in forty-one patients were reviewed retrospectively. The average duration of follow-up was six years (range, two to thirteen years). At the time of the latest follow-up evaluation, thirty-eight patients were able to perform activities of daily living, one had a stiff elbow because of heterotopic ossification, one had weakness secondary to an injury of the radial nerve, and one had an unstable elbow after removal of the prosthesis because of recurrent aseptic loosening. Fourteen patients sustained either a fracture or a perforation of the cortex at the time of removal of the primary implant. Three of these patients had an injury of the radial nerve; the injury was due to extravasation of the cement from a cortical defect in two of them and was sustained during removal of the cement in one. Eight patients had an intraoperative or postoperative complication that necessitated additional operative intervention. Postoperatively, twenty-two patients had complete relief of pain and sixteen had mild discomfort. Three patients remained disabled: one, because of pain secondary to loosening of the component; one, because of a pre-existing nerve injury; and one, because of the residual effects of an intraoperative injury of the radial nerve. The average Mayo elbow performance score was 87 +/- 16 points at the latest follow-up evaluation, compared with 44 +/- 17 points preoperatively (p < 0.0001). Revision elbow arthroplasty restored function to the patients who had had a failed prosthesis without infection.