A controlled-release table of amoxicillin trihydrate was developed by use of a matrix formulation based on the enteric polymer hydroxypropyl methylcellulose acetate succinate (HPMCAS). Sustained drug release was shown by in vitro dissolution testing; the polymer could suppress drug release in the presence of gastric pH but could enhance drug release in the presence of small intestinal pH, compared with compacts of pure drug. Grinding or physical mixing of the drug with the polymer, an alteration in normal compaction pressure, or a substitution of other enteric polymers did not markedly affect drug release from compacts. Physicochemical testing of samples confirmed that the method of mixing did not alter powder morphology. An ethanolic granulation procedure was used in the production of final tablets (21 x 10 mm) containing amoxicillin (750 mg), HPMCAS, anhydrous directly compressible lactose, and lubricants. These large tablets showed a promising sustained-release effect in vitro when a variable-pH-shift dissolution procedure was used. However, single-dose studies with a panel of fasting subjects showed that the tablets had a relative bioavailability of only 64.4%. Other pharmacokinetic parameters confirmed a lack of therapeutic advantage of these tablets over an equivalent dose of conventional capsules.