Implants placed in soft tissue evoke a foreign body reaction. Polymeric implants having smooth surfaces, such as silicone rubber implants, develop a nonadherent fibrogranulous tissue capsule which contracts over time and stiffens. Conventional porous implants, such as those made from textiles, usually have pores larger than 20μm and they become infiltrated with inflammatory tissue. The in vivo cell reaction to polymeric surfaces having pores smaller than 10μm has not been investigated systematically. In this study the histocompatibility of materials having mean pore diameters from 0.4 to 10μm was assessed. A material available with several different defined pore sizes Versapor filter material) was tested in vivo to determine relation between pore size and qualitative tissue response. Silicone-coated samples were also tested to determine the dependence of the observed tissue response on the implant surface chemistry. Results showed nonadherent, contracting capsules around implants having pore diameters smaller than 0.5μm. Implants with pores ranging from 1.4 to 1.9μm evoked thin, tightly adherent fibrous capsules without inflammatory cells. Porosities of 3.3μm and larger became infiltrated with inflammatory tissue. Results indicate that the obsexrved tissue response is predominantly dependent on implant surface topography and that variation in implant material may have little effect. It is concluded that a defined surface topography of 1 to 2 pm appears to allow direct fibroblast attachment to the surface independent of its chemical or electrochemical nature. Attached fibro-blasts then produce a minimal connective tissue response to the implant and prevent or diminish the presence of inflammatory cells at the implant/tissue interface.