In this study four quaternized derivatives of chitosan: trimethyl chitosan (TMC), dimethylethyl chitosan (DMEC), diethylmethyl chitosan (DEMC) and triethyl chitosan (TEC) with degree of substitution of approximately 50+/-5% were synthesized and their effect on the permeability of insulin across intestinal Caco-2 monolayers was studied and compared with chitosan both in free-soluble form and in nanoparticulate systems. Transepithelial electrical resistance (TEER) studies revealed that all four chitosan derivatives in free-soluble forms were able to decrease the TEER value in the following order TMC>DMEC>DEMC=TEC>chitosan, indicating their abilities to open the tight junctions. Recovery studies on the TEER showed that the effect of the polymers on Caco-2 cell monolayer is reversible and proves the viability of cells after incubation with all polymers. A similar rank order was also observed when measuring the zeta-potentials of the various polymers in solution form. Transport studies of insulin together with the soluble polymers across Caco-2 cell layers showed the following ranking: TMC>DMEC>DEMC>TEC>chitosan which is in agreement with the strength of the cationic charge of the polymer. In comparison to the free-soluble polymers, the nanoparticles prepared by ionic gelation of the chitosan and its quaternized derivatives had much lower effect on decreasing the TEER by opening of the tight junctions. This can be explained by the reduced available amount of positive charge at the surface of the nanoparticles. In accordance with these results, the insulin loaded nanoparticles showed much less permeation across the Caco-2 cell monolayer in comparison to the free-soluble polymers. Mass balance transport studies revealed that a substantial amount of the nanoparticles has been entrapped into the Caco-2 monolayer or attached to the cell surface. It can thus be stated that while free-soluble polymers can reversibly open the tight junctions and increase the permeation of insulin, the nanoparticles had basically only a low effect on the opening of the tight junction and the paracellular transport of insulin across the Caco-2 cell monolayer. These data convincingly show that nanoparticles consisting of chitosan and its quaternary ammonium derivatives loaded with insulin are less effective in facilitating paracellular transport across Caco-2 cell monolayers than the corresponding free polymers.