The safety of steroid withdrawal in orthotopic liver transplant (OLT) recipients has been studied in a prospective trial with a comparison control group. Sixty-four recipients of ABO-compatible grafts (42 adults, 22 children) were randomized into a steroid withdrawal (SW) group and a control group. Inclusion criteria included survival > one year post-OLT and no rejection > six months after OLT. Exclusion criteria included previous graft loss secondary to rejection, > two episodes of documented rejection, patients transplanted for autoimmune hepatitis, and patients unable to receive azathioprine. Target HPLC cyclosporine levels in both groups were 100-200 ng/ml. Thirty-three patients entered the SW group and 31 the control group at a mean of 3.5 years after OLT; follow-ups were 592 and 527 days, respectively. Two patients in each group developed biopsy-proven rejection. In the SW group one patient rejected at three months, the other at nine months. Both rejection episodes resolved with only reinstitution of oral prednisone. Of the two patients who rejected in the control group (one at 7 months, one at 11 months) one required conversion to tacrolimus and the other intravenous steroids. There were no significant differences between the two groups for prednisone, azathioprine, cyclosporine doses, cyclosporine levels, liver function tests, and white blood cell counts at base line compared with 12 months. Fasting serum cholesterol in the SW group decreased from 194 +/- 44 mg/dl at baseline to 175 +/- 37 mg/dl at one year, whereas in the control group cholesterol rose from 180 +/- 48 mg/dl to 193 +/- 44 mg/dl. In pediatric patients no significant difference in age-adjusted height velocities over one year was seen between the two groups. We concluded that dual therapy with cyclosporine and azathioprine in stable long-term liver allograft recipients is not associated with an increase in rejection incidence. Prednisone withdrawal may be associated with an improvement in lipid profiles.