The solubility of lorazepam in aqueous i.v. solutions and the potential for lorazepam sorption from i.v. admixtures in flexible polyvinyl chloride (PVC) bags and tubing were determined. The solubility of triplicate samples containing 15 mg of lorazepam and 20 ml of deionized water, 5% dextrose, lactated Ringer's, or 0.9% sodium chloride injections was determined. The decrease in lorazepam concentration from a 40-micrograms/ml admixture solution in 5% dextrose injection stored in PVC bags was calculated. Duplicate samples of (1) 50 ml in 50-ml PVC bags, (2) 100 ml in 50-ml PVC bags, and (3) 100 ml in 250-ml PVC bags were laid flat to approximate surface areas of 80, 160, and 270 sq cm, respectively, to determine sorptive loss. The continuous flow sorption of lorazepam from duplicate admixtures was tested at three rates through 180- and 350-cm PVC tubing. The solubility of lorazepam in deionized water, 5% dextrose, lactated Ringer's, and 0.9% sodium chloride injections was 0.054, 0.062, 0.055, and 0.027 mg/ml, respectively. Lorazepam solubility was pH dependent. The decrease in lorazepam concentration from the admixture solution stored in PVC bags for up to 121 hours followed biexponential kinetics that account for both the sorptive loss of the drug and the increasing ratio of bag surface area to solution volume. At least 90% of the initial concentration was maintained for five and two hours when the ratio was less than 2.0 and 2.7-2.8 sq cm/ml, respectively. The admixture solution retained a minimum of 95% of initial concentration when 50-ml aliquots were delivered at 600, 200, and 100 ml/hr through both sizes of PVC tubing. Because of its adequate aqueous solubility and slow sorption by PVC delivery components, lorazepam is suited for dilution in i.v. admixtures for treating conditions with intermittent or continuous infusions.