Immunoisolation devices consist of semipermeable membranes chosen to protect the islets from the immune system but still allow sufficient passage of nutrients, oxygen, and the therapeutic products, insulin. The exchange between the device and the microcirculation will influence the survival of the graft as well as the metabolic efficacy of the islet implant. Glucose is the important trigger factor for insulin secretion. In this study, we evaluate the in vivo glucose permeability of the Theracyte immunoisolation device at various times after implantation. Empty devices were implanted s.c. in rats. The glucose kinetics in the device was compared to that in the SC tissue during i.v. glucose tolerance tests (IVGTTs), using the microdialysis technique. In rats studied on day 1, or 1, 2, and 4 weeks after implantation, the peak glucose levels (Cmax) were significantly lower, the times-to-peak (TTP) were significantly longer, and the areas under the curve during the first 40 min (AUC(0-40)) were significantly smaller in the device than in the SC fat. However, at 3 months all parameters improved and Cmax, TTP, and AUC(0-40) in the device did not differ significantly from those measured in the SC fat. Thus, during the first 4 weeks the device constitutes a significant diffusion barrier, but at 3 months the exchange between the lumen of devices and the blood stream improves. Our data indicate that implantation of the device several months before transplantation of the cellular graft would improve the exchange across the membrane during the early posttransplant period. This should have positive effects on graft survival and function. We also suggest that microdialysis is a useful tool for evaluating the in vivo performance of macroencapsulation devices.