We reviewed 20 revision ankle fusions performed using internal compression arthrodesis with screw fixation. Clinical, functional, and radiographic results were measured at an average follow-up of 30 months (range, 12-50 months). The reasons for the index procedures were nonunion in 11, malunion in 7, infected nonunion in 1, and nonunion associated with avascular necrosis of the talus in 1 case. Fusion occurred in 15 of 20 patients. Two additional patients obtained fusion after subsequent procedures, for a final union rate of 85%. The average time to fusion was 6 months (range, 2-32 months). Nineteen additional operations were necessary in 12 patients, including three amputations for chronic infection (two infected nonunions and one chronic osteomyelitis). All but one patient had a plantigrade limb at follow-up. Seventeen of 20 patients were satisfied with their ultimate outcome, including all three patients with amputations. The three dissatisfied patients were bothered by chronic pain. Revision ankle fusion for nonunion or malunion using internal compression arthrodesis with screw fixation is beneficial for most patients. It is a technically demanding procedure that is associated with a high complication rate. Many patients can be expected to have residual pain. We emphasize the need for accurate alignment and early, aggressive treatment of infectious complications. Amputation should be considered a viable option to improve functional outcome in patients with solid, well-aligned fusions who are disabled by severe chronic pain.