Diclofenac sodium is an ideal candidate for incorporation in a controlled release device to diminish its adverse effects after oral administration. Microspheres were prepared by using sodium alginate as a polymer and CaCl2 as a cross-linking agent. In this investigation, 3(3) full factorial design was used to investigate the joint influence of the three variables: the stirring speed (X1), concentration of CaCl2 (X2) and % of heavy liquid paraffin in a blend of heavy and light liquid paraffin in the dispersion medium (X3) on the time of 80% drug dissolution (t80). Potential variables such as concentration of sodium alginate and drug: sodium alginate ratio were kept constant in the experimental design. A statistical model with significant interaction terms is derived to predict t80. The results of multiple linear regression analysis and F-statistics revealed that for obtaining controlled drug release, the microspheres should be prepared using relatively lower stirring speed, higher concentration of CaCl2 and higher percentage of heavy liquid paraffin in the dispersion medium. The X1X2 and X2X3 interactions were found to be statistically significant in nature. A response surface plot is presented to show the effects of X1, X2 and X3 on t80. The drug was released by diffusion of anomalous type. A model was validated for accurate prediction of drug release profile. Acceptable batches were identified in the experimental design with constraints on percentage drug released in 1, 6 and 8 h.