A study was designed to determine if admixtures with small volumes of four intravenous fluids (0.45% sodium chloride, 0.9% sodium chloride, 5% dextrose in water and lactated Ringer's) maintain their phenytoin concentrations over a suitable period of time to allow intravenous infusion of the drug. Three phenytoin concentrations (4.6 mg/ml, 9.2 mg/ml and 18.4 mg/ml) were prepared by adding a sufficient volume of an i.v. fluid to the appropriate volume (10.0, 20.0 and 40.0 ml) of phenytoin sodium injection (46 mg of phenytoin acid/ml) to produce a total volume of 100 ml. Each admixture was visually inspected for crystallization, the pH of each solution was determined, and the solutions were filtered through a 0.22-micrometer micropore filter. Unfiltered and filtered aliquots of all solutions were collected at 0, 0.25, 0.5, 1, 4, 8 and 24 hours following admixture. Phenytoin concentrations did not decline systematically during this period nor was there a significant difference (p greater than 0.05) between unfiltered and filtered aliquots in any of the solutions studied, except for the 9.2 mg/ml concentration of 5% dextrose in water. The phenytoin concentrations in the 5% dextrose in water and lactated Ringer's solutions showed the greatest variability over the 24-hour period. The pH for all solutions ranged from 10.15 to 11.50. One-half normal saline and normal saline in small volumes appear to be suitable vehicles for intravenous infusion of phenytoin.