The interlot variability in apparent pH of injectable phenytoin sodium products and the compatibility and stability of phenytoin sodium in admixtures of these products in 0.9% sodium chloride injection were studied. Dilantin (Parke-Davis) and three generic products (Elkins-Sinn, Lyphomed, Solopak) were used. Six lots of each product were diluted to 9.2 and 18.4 mg/mL concentrations. The apparent pH values of undiluted lots and diluted product admixtures were measured. Phenytoin concentrations in the admixtures were measured by turbidimetric immunoassay. Concentration and pH were measured immediately after admixture and at 0.5, 1, and 2 hours; samples were examined for crystallization at each time, and portions were filtered to determine differences in drug concentration between the filtered and unfiltered samples. The Dilantin lots had the lowest interlot variability and significantly higher mean +/- S.D. apparent pH (12.00 +/- 0.06) than the Solopak (11.38 +/- 0.33), Elkins-Sinn (11.39 +/- 0.21), and Lyphomed (11.68 +/- 0.36) products. The apparent admixture pH was significantly higher for Dilantin than for the other products. No crystallization occurred in the Dilantin admixtures; crystallization varied in the other products. Filtration did not significantly alter phenytoin concentrations. No significant differences were detected in phenytoin concentrations between products or sampling times. Interlot variability in pH was lowest for Dilantin, and apparent pH values of the undiluted products and in the admixtures at both drug concentrations were significantly higher for Dilantin than for the other products. Microscopic evidence of physical incompatibility was noted in some generic product lots with lower apparent pH values. Stability of phenytoin in the admixtures over two hours was demonstrated for all four phenytoin sodium injectable products studied.