The delivery of paraldehyde in 5% dextrose injection and 0.9% sodium chloride injection was studied, and the potential interaction between paraldehyde and plastic i.v. containers and sets was evaluated. Paraldehyde was mixed with either 5% dextrose injection or 0.9% sodium chloride injection in polyvinyl chloride (PVC) bags to form a 4% solution. The bags were fitted with standard i.v. administration sets or burettes with administration sets. The solutions were allowed to drip through the i.v. sets for six hours at room temperature. Samples were taken from the i.v. bag or burette and from the distal end of the i.v. sets at zero, two, four, and six hours. Paraldehyde concentrations were measured using a stbility-indicating gas chromatographic method, and the presence of plasticizers was detected by a scanning ultraviolet spectophotometer. The cumulative amount of paraldehyde delivered at the end ot the administration set at six hours was 84% for 5% dextrose solutions in burettes, and 89% or 90% for all other solutions and i.v. sets. An ultraviolet-light-absorbing substance appeared in some of the samples, although a relationship between the presence of this substance and type of solution, time of sampling or site of sample did not emerge. Particulate matter appeared after two hours in all burettes. Approximately 10%-16% of paraldehyde in 5% dextrose of 0.9% sodium chloride injection is lost when delivered from PVC i.v. bags through standard i.v. administration sets and burettes over a six-hour period. If paraldehyde is to be administered intravenously using PVC i.v. equipment, the risk of adsorption to this equipment and potential toxicity of dissolved plasticizers must be weighed against the necessity of administering this drug to a particular patient.