 The Kosovovence models presented from 1960 to 2007 were reviewed and their accuracy for correlating and or predicting the solubility of drugs in water-cosolvent mixtures was assessed. These models can be categorized into three groups, theoretical, semi-empirical, and empirical. The simplest Kosovovence model, the log linear model of Yakovsky, estimates drug solubility in water-cosolvent mixtures by using the aqueous solubility of the drug. The Jui-Banakri model requires only one additional data point, the solubility in neat-cosolvent, to make predictions with an acceptable error rate. Error terms used in the literature were also discussed, along with some general comments about the acceptable prediction error for pharmaceutical applications. This article was authored by Abogas M. Jui-Ban.