The yield of pellets obtained in preparing diclofenac sodium (DC-Na) controlled release pellets was affected by the lactose/starch ratio of the feed granules, viscosity of the binding solution and residence time in the rolling pot. The pellet shape was found to be correlated with surface tension of the binding solution. The concentration and composition of the coating solution and pellet size distribution were responsible for the agglutination in coating. The controlled release pellets (CRP) and commercial tablets (CT) of DC-Na were administrated orally to healthy volunteers by multiple-dose. The steady-state condition, based on trough plasma concentration on day 3-5, was achieved on day 3. A great difference in plasma drug concentration fluctuation index (FI) between CRP (0.476 +/- 0.0484) and CT (0.935 +/- 0.092) was observed (P < 0.05) during steady-state. The area under the plasma drug concentration-time curve for a 0-12 h interval (AUC) on day 5 of CRP (6.493 +/- 0.4169 micrograms.h.ml-1) and CT (7.551 +/- 0.4745 micrograms.h.ml-1) was shown to be not significantly different (P > 0.05). The relative bioavailability in the human was about 86%.