A series of experiments was undertaken to evaluate the diffusion of a model protein, i.e. bovine serum albumin (BSA), from chitosan-alginate microcapsules produced using either internal or external gelation. Diffusion of BSA was quantified during the microcapsule manufacture processes (gelation, washing, rinsing) and during incubation in conditions simulating the pH encountered during the gastric and intestinal phases of digestion. Encapsulation of an acid phosphmonoesterase permitted in situ protein localization, providing evidence to explain results obtained with BSA. There was significantly greater protein loss from internally versus externally-gelled chitosan-alginate microcapsules during the manufacture process (37.6% versus 4.7%, respectively). Similar trends were observed during 24 h incubation in 0.1 N hydrochloric acid. Increasing alginate concentration from 2-4% (w:v) did not significantly reduce losses from internally-gelled microcapsules. Addition of 0.25 M NaCl to the gelling medium significantly increased protein diffusing during microcapsule manufacture and acid incubation from externally gelled microcapsules. In situ protein localization revealed a higher level of protein near the surface of the microcapsules of externally gelled microcapsules versus internal gelation. The above data indicate that externally-gelled microcapsules are inhomogeneous with a higher concentration of alginate near the microcapsule surface, thus reducing the porosity of the resulting microcapsules. These results suggest that the porous nature of internally-gelled chitosan-alginate microcapsules may result in low encapsulation efficiency, depending on the nature of the product being encapsulated.