This study assesses whether in vitro immediate release ketorolac tablet dissolution profiles (utilizing the recently proposed USP dissolution test for ketorolac tablets) can be correlated with in vivo plasma pharmacokinetic parameters. Four batches of ketorolac tablets were utilized: a ketorolac tablet batch that demonstrated a rapid dissolution rate during USP in vitro dissolution testing, two tablet batches that were manufactured such that they dissolved at moderate rates, and a tablet batch that was manufactured such that it dissolved at a distinctly slow rate. The single-dose mean pharmacokinetic characteristics and relative bioavailability of the four different 10 mg ketorolac tromethamine tablets were evaluated in 12 healthy volunteers in a randomized study of Latin square design. The amount dissolved of the various tablets at 10, 20, and 30min was in the order of fast-dissolving tablets > medium-1 -dissolving tablets = medium-2-dissolving tablets > slow-dissolving tablets. In general, the profiles of the average plasma concentrations for ketorolac were similar for the fast-and the two medium-dissolving tablet batches (even though a statistically significant difference was found between the tmax of the fastdissolving tablet and one of the medium-dissolving tablet batches). The mean plasma concentrations for the slow-dissolving tablet, however, reached peak levels much later, with the peak also being significantly smaller. There were no statistically significant differences in the total AUC or in the mean plasma halflives among the four formulations. Good correlations were obtained for mean t,,, versus the percentage dissolved at 20, 30, and 45min. Correlations were generally weaker for percentage dissolved versus C,,, or percentage bioavailability. This indicates that in vitro dissolution testing for immediate release ketorolac tablets can be a useful indicator of in vivo time to maximum plasma concentration when comparing similarly formulated tablets. Further, the proposed USP dissolution test and specification would have appropriately failed the slow-dissolving tablet batch, which demonstrated a significantly slower rate of absorption as per tmax and Cmax.