The purpose of this study was to evaluate the potential utility of chitosan (I) in inhibiting diclofenac sodium (II) release in the gastric environment from a directly compressible tablet formulation. I, subjected to depolymerization to improve its microcrystallinity and subsequent compressibility, was then used to prepare tablets of II. A full-factorial design was employed to evaluate the effects of degree of N-deacetylation of I, and the pH and ionic strengths, mu, of the dissolution media on drug release. Directly compressible tablets were prepared from admixtures of 25 mg of II, 174 mg of I of various degrees of N-deacetylation (74, 87, and 92%), and 1 mg of magnesium stearate. The in vitro dissolution studies were performed using aqueous buffers (pHs 1.2, 3.8, and 6.8, and mu of approximately 1.0 and 0.1). The slopes of logarithmically transformed cumulative percent released-time curves (from t = 0 to t = 5 hr) were compared. Analyses of variance performed using SAS indicated that the degree of N-deacetylation of chitosan significantly affected drug release at pHs 1.2 and 6.8 (p < 0.0001). An increase in the pH of the dissolution medium resulted in an increase in drug release (p < 0.0001). The ionic strength of the dissolution medium did not significantly affect drug release at any of the pHs studied (p > 0.198). Besides the poor aqueous solubility of II, the two factors possibly affecting the drug release in the acidic environment were (a) the formation of a rate-limiting chitosan gel barrier; and (b) the ionic interaction of II with ionized amino groups of I.