Indomethacin (IDM) was encapsulated in gelatin-cellulose acetate phthalate (CAP) microcapsules (A) by complex coacervation method and in CAP microcapsules (B) by simple coacervation method. Microcapsules A and B, having mean diameters of 38.24 and 35.74 μm, respectively, were used to prepare sustained-release tablets A and B. The activation energy of thermal degradation for tablets A and B was calculated based on differential scanning calorimetry (DSC) to be 258.9 and 284.8 kcal/mol, respectively. In vitro release profiles showed no burst effect and release t 1/2 of the two sustained-release tablets were found to be 41.30 ± 1.86 and 33.25 ± 2.84 min, respectively, while that of IDM plain tablets C was 6.30 ± 0.39 min ( P < 0.01). In vitro release of IDM from tablets A and B could be described by Higuchi equation and zero-order kinetics, respectively. After per os (po) administration with physiological saline, their irritation to rat stomach was obviously reduced in comparison with tablets C. Pharmacokinetic study in rabbits showed that t max was delayed and C max lowered compared with tablets C and the values of AUC 0–24 h of the three tablets were very close.