Foreign-body (FB) carcinogenesis is a classic model of multistage tumour development in rodents. Previous studies have demonstrated that the physical characteristics of the implant, and not the chemical composition, are the critical determinants of tumour development. The recent controversy over silicone breast implants has raised questions regarding the potential carcinogenicity of lifetime tissue exposure to silicone products. The present study was designed to determine whether the inflammatory and fibrotic reactions associated with silicone implants are due to a non-specific foreign-body reaction or whether these responses reflect the unique chemical composition of silicone. F344 rats were implanted subcutaneously with one of three biomaterials: silicone elastomer (Group 1); impermeable cellulose acetate filters (Group 2, positive control); or porous cellulose acetate filters (Group 3, negative control). The silicone and cellulose implants of Groups 1 and 2 have been previously shown to induce fibrosarcomas in rodents, whereas the porous cellulose acetate implants of Group 3 have been shown to be non-carcinogenic. One week and two months after implantation, the pericapsular tissues were evaluated using histopathological and in situ immunohistochemical analyses. Endpoints included expression of leucocyte antigens CD4 (T helper/inducer), CD8 (T suppressor/cytotoxic) and CD11 b/c (macrophage), proliferating cell nuclear antigen (PCNA) as an indicator of proliferation, and in situ end-labelling (ISEL) of 3'OH DNA strand breaks as an indicator of DNA damage and apoptosis. The results indicated that the acute and chronic cellular responses to silicone (Group 1) were not different from impermeable cellulose filters (Group 2) of identical size and shape, suggesting that these responses were not unique to silicone. The inflammatory response to the carcinogenic cellulose and silicone implants (Groups 1 and 2) was attenuated and associated with the formation of a thick fibrotic capsule. In contrast, the porous cellulose filters (Group 3) induced a markedly different cellular response in which the inflammatory reaction was more extensive, prolonged and associated with minimal fibrosis. Within the fibrotic capsule surrounding the tumorigenic implants, but not the non-tumorigenic implants, cell proliferation and apoptotic cell death were increased and associated with persistent DNA strand breaks. Taken together, the results suggest that the micrometre-scale surface morphology of the implant determines the nature of the subsequent cellular response which may predispose to tumour development. Further, these studies serve to emphasize the critical importance of appropriate physical controls in studies designed to evaluate carcinogenic or autoimmune manifestations associated with silicone implants in order to rule out the contribution of the chronic foreign-body reaction.