The effect of rapamycin (RAPA) on both host-versus-graft (HVG) and graft-versus-host (GVH) immune responses was examined in small bowel transplant models using strongly histoincompatible donor-recipient combinations. Normal Wistar Furth (WFu; RT-1u) recipients rejected Buffalo (BUF; RT-1b) small bowel allografts within a mean survival time (MST) of 10.5 +/- 0.5 days. Administration of RAPA (0.8 mg/kg) by continuous intravenous infusion for 14 days via an osmotic pump prolonged graft survival to 25.0 +/- 4.6 days (P = 0.01). In a second strain combination, the 12.5 +/- 2.2 day survival of Brown Norway (BN; RT-1n) small bowel allografts in Lewis (RT-1l) recipients was prolonged to 21.6 +/- 2.0 and 28.5 +/- 2.8 days by 14 days of i.v. RAPA at doses of 0.8 and 1.6 mg/kg, respectively. In this model RAPA is five times more effective than cyclosporine, which at 4.0 mg/kg prolongs BN small bowel allografts in Lewis recipients to 21.6 +/- 6.3. To isolate HVG and GVH immune responses, (BN x Lewis)F1 hybrid rats served as the graft donor or host, respectively. In the HVG model, (BN x Lewis)F1 small bowel allografts, which were rejected by normal Lewis recipients at 12.2 +/- 3.6 days, were prolonged to 40.8 +/- 5.8 days (P = 0.001) by RAPA (0.8 mg/kg x 14 days). In the GVH model, the ability of Lewis small bowel allografts to produce severe GVH disease in untreated (BN x Lewis)F1 recipients at 12.3 +/- 2.8 days was delayed to 21.3 +/- 5.2 days by 0.8 mg/kg RAPA (P = 0.025). Thus, RAPA protects small bowel allografts more effectively against HVG than GVH immune responses.