We performed a 1-year study to determine whether intermittent short courses of the microemulsion formulation of cyclosporin (Neoral) could effectively control plaque psoriasis and whether tapering or abrupt cessation of cyclosporin therapy would influence time to relapse. Four hundred patients with plaque psoriasis were included in this open, multicentre, randomized study. All patients commenced cyclosporin at a dose of 2.5 mg/kg daily. Cyclosporin dosage could be increased to a maximum of 5 mg/kg daily. Treatment was continued until clearance of psoriasis or for a maximum of 12 weeks. Patients were then randomly assigned either to stop cyclosporin abruptly or to have the dose reduced by 1 mg/kg daily each week until cessation. On relapse, patients were given another course of cyclosporin. Patients were followed for at least 1 year, during which they could receive as many treatment courses as necessary. The number of patients who required one, two, three and four treatment courses was 400, 259, 117 and 26, respectively. The median time to relapse after the end of the first treatment period was 109 days in the group of patients randomized to stop cyclosporin abruptly and 113 days in patients randomized to taper off cyclosporin (P = 0.038). More than 30% of patients had not relapsed 6 months after having stopped treatment. After each treatment course, the Kaplan-Meier probability of achieving 75% or more reduction in disease area by day 84 of treatment was 83%, 76%, 73% and 66%, respectively. Mean serum creatinine concentration and blood pressure did not show any clinically significant changes over time. Our results show that intermittent short-course therapy with Neoral, when used in conjunction with topical therapy, is well tolerated and provides effective control of plaque psoriasis for 1 year. Tapering off cyclosporin on treatment cessation induces a slight delay in psoriasis relapse.