Acute cellular rejection of the allograft is a potentially serious complication after liver transplantation, yet its true incidence is unknown. We therefore investigated the frequency of acute cellular rejection reported by transplant centers and its impact on morbidity and mortality. Morbidity was defined as duration of hospitalization. Of 200 articles screened, 18 were selected for inclusion in the study database, in which there was a total of 1,437 patients who received transplants. All contained more than 20 patients and invariably used histopathology for diagnosis of acute cellular rejection. These reports included all transplant patients within a fixed period and sufficient data to determine the incidence of acute cellular rejection. Morbidity data were obtained from our previous series. The mean incidence of acute cellular rejection in all centers was 49.8% (range between centers, 24% to 80%). Two immunosuppressive cohorts were identified: high-dose cyclosporine induction (> or = 5 mg/kg/d) and low-dose cyclosporine induction (< or = 4 mg/kg/d). Acute cellular rejection was reported in 27.0% of the high-dose group and 63.6% of the low-dose group, P = .0001. Strict adherence to Snover's histological criteria for acute cellular rejection did not alter the reported mean incidence. Frequency of acute cellular rejection was 45.2% (range between centers, 24% to 80%) in 8 studies that used Snover's criteria, and 51.6% (range between centers, 37% to 80%) in 10 studies that did not. There was no correlation between mortality and incidence of acute cellular rejection in the 9 studies that reported survival (R2 = .105). Morbidity data showed that the average length of initial hospitalization after transplantation for patients with acute cellular rejection was 52.4 +/- 8.3 (range, 14 to 124) days, in contrast to 28.3 +/- 2.3 (range, 9 to 87) days for patients with no rejection. P = .0008. The total number of hospital days in the first 6 months for patients with acute cellular rejection was 55.6 +/- 8.6 (range, 14 to 124) days and with no rejection, was 37.7 +/- 3.1 (range, 9 to 99) days. P = .0232. The incidence of acute cellular rejection varies widely among transplant centers, regardless of the use of Snover's criteria. Acute cellular rejection appeared to be less frequent in programs using high-dose cyclosporine induction regimens. The presence of acute cellular rejection seemed to have no correlation with mortality but significantly increased morbidity and therefore the cost of transplantation.