This study assesses the plasma-tissue exchange characteristics of the capsular tissue that forms around implants and how they are affected by implant porosity. The number of vessels and their permeability to rhodamine were measured by intravascular injection of the fluorophore tracer into Sprague-Dawley rats that hosted for 3-4 months polyvinyl alcohol (PVA) and polytetrafluoroethylene (PTFE) subcutaneous implants. Rats were implanted with four pore sizes of PVA-a nonporous PVA (PVA-skin), and 5, 60, and 700 micron mean pore sizes (PVA-5, PVA-60, and PVA-700, respectively)-and two pore sizes of PTFE: 0.50 (PTFE-0.5) and 5.0 (PTFE-5) mean micron pore sizes. Photodensitometric image analysis was used to quantify the local tracer extravasation and, hence the permeability coefficients of isolated vessels around the implants. The number of functional vessels within 100 m of the implants highlighted by the lissamine-rhodamine tracer were counted with fluorescence microscopy and with H&E stained sections using brightfield microscopy. The permeability of vessels did not vary substantially with implant pore size but generally were lower than those measured for surrounding subcutis. Pore size, however, had a dramatic effect on the vascular density of tissue-encapsulating implants: the number of microvessels (under 10 m in radius) within the tissue surrounding the porous implants was higher than the number around nonporous implants. Pore sizes on the order of cellular dimensions incited optimal neovascularization; the vascular density around PVA-60 implants was six times higher (p < .001) and three times higher (p < .001) than those around PVA-0 implants in the fluorescent images and in brightfield, respectively. Moreover, brightfield microscopy showed the number of vessels around PVA-60 implants was almost double those in normal subcutis. The results suggest that optimal vascular density around long-term implants, such as sensors, biofluid cell constructs, and immunoisolated cell systems, may be engineered with pore size.