We describe a computational model and an in vitro experiment for assessing whether or not a pH-solubilized drug has the potential to precipitate upon dilution with blood. The computational model enables an efficient means of selecting buffer concentration and pH, and the in vitro test provides a simple experimental validation. Both means of screening are applied to the formulation of the weakly basic drug levemopamil-HCI. A buffered formulation of levemopamil is chosen from the computational model and shown to be free of precipitation upon dilution in vitro and to not produce phlebitis in the rabbit ear model. In comparison, an unbuffered formulation at the same pH and drug concentration precipitates in vitro and causes significant phlebitis in vivo. The results of this study reinforce the importance of buffering parenteral formulations instead of simply adjusting the pH of the formulation.