The purpose of this investigation was to determine whether potassium chloride as an additive is distributed homogeneously throughout small volume parenteral solutions and whether the pattern of electrolyte distribution is identical for two types of delivery systems. The effect of purposeful mixing after addition of potassium chloride was also studied. Thirty-one registered nurses prepared two supplementary intravenous solutions consisting of 10mEq potassium chloride in 50ml of 5% dextrose solution. One solution was prepared in a soft polyvinyl chloride minibag (Viaflex®, Baxter-Travenol), the other in a semirigid plastic burette (Buretrol®, Baxter-Travenol). The solutions were discharged through a volumetric pump (IMED®) and four aliquots were taken at regular intervals. Samples were chemically analyzed by flame photometry to determine the potassium concentration in each solution layer of the container. For each container type, there were significant differences in the potassium concentrations among the layers of solution. The pattern of distribution also differed significantly between the two containers. The strata in the minibag showed smaller variations in potassium concentration than did corresponding layers in the burette. Some single sample concentrations from burettes exceeded the expected concentration of 200mEq/L by 105mEq. Nurses manipulated the two types of containers differently. Thirty of 31 minibags were mechanically mixed, but only 11 of 31 burettes were treated in this way. When the effect of mixing was considered, solutions in mixed containers were essentially no different from those not agitated after addition of the electrolyte.