We compared results using Neoral versus Sandimmune, each in combination with steroid and azathioprine immunosuppression, in primary liver transplantation recipients. There were 15 patients in each group with similar demographic distributions. Intravenous cyclosporine was stopped at 4.3 +/- 1.9 days in the Neoral group vs 7.8 +/- 4.9 days in the Sandimmune group. (P < 0.025). Cyclosporine levels in the first 10 days were higher (mean 306 ng/ml vs 231 ng/ml) in the Neoral group than the Sandimmune group (P < 0.05). The Neoral dose was less than the Sandimmune dose (mean 5.5 ng/kg per day vs 7.9 ng/kg per day) to achieve these levels in that time period (P < 0.05). Two patients (13%) experienced three episodes of biopsy-proven rejection in the Neoral group compared to nine patients (60%) with 12 episodes of rejection in the Sandimmune group (P < 0.025). Incidences of neurological and renal complications were similar between the groups. Infections requiring treatment were also similar. Liver function, renal function, and marrow function, evaluated at days 7, 14, 21, 28, and 2, 4, 6, and 12 months post-transplant, were not different between the groups. In summary, shorter use of intravenous cyclosporine and quicker stabilization of trough cyclosporine levels was achieved with Neoral than with Sandimmune. In the early post-transplant period, higher levels with lower doses were achieved with Neoral than with Sandimmune. In our experience, the incidence of rejection was lower with Neoral than with Sandimmune. There were similar lengths of hospitalization, mortality, adverse events, retransplantation, and similar liver, renal, and marrow function up to 1 year post-transplantation. Because of this experience, we continued to use Neoral in a total of 59 primary liver transplant recipients. We have not used intravenous cyclosporine in the last 44 patients. Follow-up was a mean of 11.4 months, ranging from 1 to 27 months. The incidence of rejection was 24% in these 59 patients compared to our historical experience of 70% using Sandimmune.