In this retrospective study, we evaluated the histological and biological predictors of long-term response of renal transplant (RT) patients treated with orthoclone OKT3 for steroid resistant acute rejection (AR). Seventy-three patients, aged 37 +/- 12 years, were included in this study between March 1987 and December 1996. All the patients but one had received sequential quadruple immunosuppression (polyclonal antilymphocyte globulins; steroids; azathioprine, and cyclosporin A). OKT3 (5 mg/day for 10 days) was administered for biopsy-proven steroid resistant AR i.e., after 3 consecutive pulses of methylprednisolone (10 mg/kg each). This was the first AR in 46 cases, the second AR in 22 cases and the third AR in 4 cases. Renal histology (Banff) showed borderline (BL) changes in 18 patients, grade I AR in 28 patients; grade II AR in 22 patients, and grade III AR in 5 patients. When treatment with OKT3 commenced (107 +/- 18 days post-transplantation) the mean serum creatinine (SCr) level was 325 +/- 195 micromol/l; this had decreased to 191 +/- 106 micromol/l by the end of OKT3 therapy. The immediate response to OKT3 therapy i. e., within the first month, was not dependent on the histological score. Twenty-six patients (35%) subsequently experienced at least one more AR episode of whom 4 were retreated with OKT3. The overall patient's survival was 94.5% at last follow-up. The overall cumulative graft survival was 64.5% at 2 years, 52.5% at 5 years, and 40.5% at 8 years. The graft survival (5 years) tended to depend on the initial histological score, i.e. BL 30%; grade I 66%; grades II and III 55.5% (p = 0.08). In a multiple logistic regression analysis we tried to identify independent factors that would predict that a graft would still be functioning at least 2 years after OKT3 therapy. We therefore analyzed the following parameters: donor and recipient's age; gender; cold ischemia time; HLA matching; panel reactive antibodies (PRA) prior to grafting; previous transplantation(s); total number of AR episodes; the time of onset of the AR treated by OKT3 compared to the other AR; the time of onset of the AR treated by OKT3; SCr levels at days 0, 10 and 30 after OKT3 therapy; histological score (Banff) i.e., the magnitude of AR and the presence or absence of chronic lesions. The only independent factors which would predict that a graft was still functioning 2 years after OKT3 therapy were: PRA <25% (Odds ratio (OR) 7.68 (1.15-51.3); p = 0.035); a grade I AR (OR 10.52 (1.18-93. 5); p = 0.035); SCr level 1 month after OKT3 therapy (OR 0.935 (0. 87-1.002); p = 0.05). HLA matching and the presence of histological chronic lesions were nearly significant (p = 0.06 and 0.09 respectively). In conclusion, this retrospective study shows that independent predictors of the long-term response to OKT3 therapy for AR in RT patients are the magnitude of pre-transplant PRA, the histological score, and the SCr level one month after OKT3 therapy.