The sorption of diazepam in large-volume i.v. admixtures to administration-set components and in i.v. containers was analyzed quantitatively. Solubility of diazepam in phosphate buffer at various pH levels and in i.v. fluids was measured. Partition coefficients of diazepam into components of i.v. administration sets and i.v. containers were studied by shaking a solution of diazepam in 0.9% sodium chloride, with finely cut components and measuring the change in diazepam in the aqueous phase. Flow studies through an administration set of a 0.04-mg/ml diazepam solution in 5% dextrose injection were done, varying both the flow rate and the length of tubing. The maximum free-base solubility of diazepam in phosphate buffer was 0.048 mg/ml; its solubility was 0.058, 0.050, and 0.064 mg/ml in lactated Ringer's, 0.9% sodium chloride, and 5% dextrose injections, respectively. Equilibrium partition coefficients were highest for polyvinyl chloride tubing and flexible bags. Volume-control sets made of cellulose propionate had lower but sufficiently large partition coefficients to cause diazepam loss. Polyolefin semi-rigid and glass containers had low partition coefficients. In the flow studies, the amount of solution-contact time correlated with the extent of absorption. As flow rate decreased or tubing length increased, the amount of diazepam absorbed increased proportionately. A nomogram and a predictive dosing chart are presented for calculation of actual diazepam doses delivered at various flow rates and tubing lengths. Diazepam can be administered safely and effectively by i.v. infusion. The use of volume-control sets and flexible polyvinyl chloride bags should be avoided with diazepam solutions. Polyolefin semi-rigid containers are acceptable alternatives to glass. The concentration of diazepam infusions should not exceed 0.04 mg/ml.