To assess the safety profile of Neoral dose adjustment using cyclosporine (CsA) trough levels (C0) compared with levels obtained 2 h after the morning dose (C2), 30 stable adult heart transplant patients 1 yr or more after surgery were converted from Sandimmune to Neoral. After a baseline visit (before conversion), initial follow-up included two visits (2 and 4-6 wk after conversion). After the first visit, patients were randomized to Group I (C0: 100-200 ng/ml) or Group II (C2: 200-400 ng/ml). Abbreviated pharmacokinetics were obtained for the estimation of the AUC0-4 h. Renal function was assessed by serum creatinine and the cimetidine-modified creatinine clearance. C2 correlated better than C0 with the AUC0-4 h (r = 0.91 vs. 0.63). Initial Neoral dose (mg/kg/d) was similar in both groups (2.8 +/- 0.5 and 2.8 +/- 0.8), and was lower in Group II at the second visit (2.0 +/- 0.7 vs. 3.0 +/- 0.6, p = 0.0001). C2 levels decreased in Group II from 912 +/- 438 to 555 +/- 271 ng/ml (p = 0.01), without evidence of acute rejection on endomyocardial biopsies. After the second visit,-both groups were monitored with C2, and the range was increased to 300-600 ng/ml. At the last visit (additional follow-up of 5 +/- 1 months), Neoral dose (mg/kg/d) was reduced to 2.0 +/- 0.3 in Group I (p < 0.001) and 1.8 +/- 0.4 in Group II. Serum creatinine was lower in Group II at the second visit (138 +/- 59 vs. 168 +/- 37 mumol/L, p = 0.01) and was similar in both groups at the last visit. Neoral dose reduction based on C2 levels was not associated with acute rejection. The better correlation with the AUC0-4 h suggests that C2 may be more reliable than C0 for Neoral dose adjustment.