The objective of this work was to prepare and evaluate ketorolac tromethamine-loaded albumin microspheres using a factorial design. Albumin microspheres were prepared by emulsion cross-linking method. Selected formulations were characterized for their entrapment efficiency, particle size, surface morphology, and release behavior. Analysis of variance (ANOVA) for entrapment efficiency indicated that entrapment efficiency is best fitted to a response surface linear model. From the statistical analysis it was observed that as the drug:polymer (D:P) ratio and volume of glutaraldehyde increased, there was a significant increase in the encapsulation efficiency. Scanning electron microscopy of the microspheres revealed a spherical, nonporous and uniform appearance, with a smooth surface. Based on the entrapment efficiency and physical appearance, 9 formulations were selected for release study. The maximum particle size observed was below 40 microm. The release pattern was biphasic, characterized by an initial burst effect followed by a slow release. All selected microspheres, except those having less polymer proportion (D:P ratio is 1:1), exhibited a prolonged release for almost 24 hours. On comparing r (2) values for Higuchi and Peppas kinetic models, different batches of microspheres showed Fickian, non-Fickian, and diffusion kinetics. The release mechanism was regulated by D:P ratio and amount of cross-linking agent. From the experimental data obtained with respect to particle size and extent of drug release, it could be concluded that the prepared microspheres are useful for once-a-day intramuscular administration of ketorolac tromethamine.