The aim of this study was to investigate different gas-permeable membranes for culturing human islets. Dynamic insulin release was used to assess islet functional quality. Islets isolated from cadaveric pancreata (n = 8) using standard isolation methods were stained with dithizone, counted, and cultured on five different commercially available medical-grade membranes reported to have high permeability to O2, CO2, and other gases. Fraction 1 (20,000 islet equivalents [IEQ] purity >70%; viability >85%) was cultured using serum-free medium in nonadherence tissue culture flasks (group I) and custom-made chambers with membranes (group II). Each vessel contained 5000 IEQ at a density of 30 IEQ/cm2 and 69 IEQ/cm2 for groups I and II, respectively. Islets were cultured for 48 to 90 hours at 37 degrees C in 5% CO2. In vitro dynamic insulin response to low glucose (3 mmol/L), high glucose (16.7 mmol/L), and 25 mmol/L KCI was measured. Stimulation indices were calculated by dividing average of initial response over basal insulin release; basal insulin release defined as average of the first seven values. Islets cultured on MG7 (n = 3) showed a higher stimulation index (3.49 +/- 0.37) compared with flasks (2.44 +/- 0.22), indicating better specific functional quality. Islets cultured on other membranes proved to show similar or worse functional quality than those cultured in flasks. In fact, islets cultured on MG6 (n = 2) were not tested owing to complete disintegration. Islet functional quality was improved when cultured on selected biocompatible gas-permeable membranes; however, finding the best membrane requires further investigation before clinical application.