We conducted a phase II trial in 19 chronic graft-versus-host disease (cGVHD) patients with rapamycin, calcineurin inhibitors, and prednisone with the goals of controlling cGVHD, reducing prednisone use, and defining the safety of this regimen. Rapamycin was begun as second-line (n = 9) or more than second-line (n = 10) therapy. With a median follow-up of 42 months, 16 patients were evaluable for response. Nine patients discontinued rapamycin because of poor compliance/patient request (n = 2) or an adverse event (n = 7), 3 of whom were not evaluable because of withdrawal at < or =1 month or noncompliance. The adverse events included serum creatinine > or =2.4 mg/dL (n = 4), hemolytic uremic syndrome (n = 2), and relapse of malignancy (n = 1). Fifteen of 16 evaluable patients had a clinical response. Five of the 16 discontinued the drug, and 1 died of relapsed leukemia. Of the 10 patients who continued rapamycin, 2 discontinued and 1 successfully tapered all systemic immunosuppression. Three of the 10 developed progressive cGVHD with tapering immunosuppression; all responded to resumption of prior medications. Four of the 10 patients required alternate therapy for persistent or progressive cGVHD while receiving rapamycin; prednisone was discontinued (n = 2) or tapered at the time of progressive disease (n = 2). Seventeen of 19 original patients were alive. One death was due to relapsed malignancy, and 1 was due to congestive heart failure. In this report of rapamycin as cGVHD therapy, there is evidence of rapamycin's efficacy. Given the significant toxicities described, investigation of altered administration of rapamycin and calcineurin inhibitors should be pursued in future cGVHD trials.