Biofunctional membranes are entities in which biological molecules (or cells) are attached to polymeric supports cast in the form of porous membranes. Such membranes are gaining increased importance in applications of enzymatic catalysis or synthesis (bioreactors), separations (affinity membranes), and chemical analysis (biosensors). However, fundamental studies of the active site of immobilized biomolecules have been rare. In this study, electron paramagnetic resonance (EPR) spin-labeling techniques using a short, active-site specific spin label have been employed to study the properties of a model enzyme, papain, immobilized on a fully-hydrated, modified polysulfone membrane. The EPR properties of the immobilized enzyme and reaction rate results using the amidase activity of papain and N-benzoyl- dl-arginine-4-nitroanilide hydrochloride as substrate are compared with that of the free enzyme in solution. The major findings in this study are: (1) Immobilization does change the active-site conformation of papain. The spin label at the active site of the immobilized papain has slower motion than that of the free papain in solution. (2) There are two major subpopulations of immobilized enzyme on modified polysulfone membranes: subpopulation A has faster spin-label motion at the active site than subpopulation D, suggesting that the enzyme in subpopulation A may have a more open active-site cleft than that of the subpopulation D. (3) The active-site conformation of subpopulation D of the immobilized papain is insensitive to the pH of the bulk solution, while that of subpopulation A has a similar response to pH changes as that of free papain in solution, suggesting that subpopulation A may be the active form of the immobilized enzyme while subpopulation D is the denatured form. (4) The pH-dependent curve of the amidase activity of the immobilized papain has a similar bell shape as that of the free papain in solution. (5) Subpopulation A is converted into subpopulation D upon addition of denaturants (urea or guanidine hydrochloride), which further confirms the assumption that the former subpopulation is the active form of the immobilized enzyme and the latter subpopulation is the denatured form. Moreover, denaturation at elevated temperature can also convert subpopulation A into subpopulation D. (6) By paramagnetic relaxation mechanisms, which require close proximity of the paramagnetic species and the spin label, K 3Fe(CN) 6 can broaden the EPR signal of spin label bound to subpopulation A but not that of subpopulation D, suggesting that subpopulation A is accessible to the substrate. More direct evidence that subpopulation A is the active form of the immobilized enzyme and more accessible to substrate than subpopulation D was obtained by immobilizing papain on modified polysulfone membranes first followed by spin labeling. In this way, only the active subpopulation A was labeled by the spin label. (7) The K m(app) values of the immobilized papain are larger than that of papain in solution, while V max(app) is smaller. (8) The enzyme is highly reusable and has very high storage stability after immobilization. (9) Papain immobilized on polysulfone membranes has much higher thermal stability and stability toward denaturants (urea or guanidine hydrochloride) than that of free papain in solution. In this study the polymeric membrane provided three functions: sites for covalent coupling of enzymes; enhanced enzyme stability; and substrate partitioning. This paper reports the first development of an effective method to acquire EPR spectra of fully-hydrated, spin-labeled enzymes immobilized on polymeric membranes. All these findings indicate that the EPR spin-labeling technique shows great promise as a powerful method for studying membrane-immobilized enzyme systems.