The effectiveness of local versus systemic low-dose CsA (2 mg/kg/day) therapy delivered by osmotic pump for a 14-day continuous infusion was examined in the rat model. Systemic subtherapeutic CsA treatment of WFu recipients either by oral gavage or intravenously using an osmotic pump resulted in quick rejection of BUF heart allografts within a median survival time (MST) of 8 days in comparison with untreated controls (MST = 7 days). In contrast, direct local subtherapeutic CsA delivery to BUF heart allografts produced significantly (P less than 0.01) prolonged heart allograft survivals up to MST of 40 days. Splenic T cells, isolated on days 10 to 12 from locally immunosuppressed WFu recipients, revealed a nonspecifically reduced proliferative response toward alloantigens. Coculture experiments demonstrate that these T cells have the capacity to inhibit normal T cell proliferative responses in a nonspecific fashion either by their suppressor function or more likely by carrying CsA to the culture plate. In contrast, T cells isolated from WFu recipients three weeks after transplantation and tested in vitro demonstrated the presence in alloantigen specific T suppressor cells that coincided with a decreased frequency of alloantigen-specific T cytotoxic cells and may explain the extended heart allograft survival beyond the time of CsA delivery. CsA therapy delivered directly to the graft resulted in high CsA levels within the heart graft (1108 ng/0.1 g) but subtherapeutic levels in other tissues. These results demonstrate that local drug delivery is effective in inhibiting the rejection process within the graft itself, as manifested by prolonged heart allograft survival. Further, subtherapeutic CsA therapy facilitates development of Ts cells that may be responsible for the survival of heart allografts beyond the CsA delivery time.