A taste sensor instrument (electronic tongue) was evaluated to determine its utility in developing a taste-enhanced liquid formulation. To train the electronic tongue, human sensory panel data were collected for two prototype formulations, a solution of the drug in water and several marketed products. Studies using the electronic tongue were conducted to determine taste-masking effectiveness of formulations compared to a matching placebo, to establish correlation with human sensory data, and to evaluate unknown formulations and predict their bitterness scores. In the first experiment, the effectiveness of a proposed taste-masking strategy was determined by comparing formulation prototypes containing a bitter active pharmaceutical ingredient (API) against corresponding placebos (i.e. formulations without an active ingredient) using electronic tongue data. The analysis of the electronic tongue data was based on the assumption that the drug was well taste masked if the placebo matched the formulation with API. In a second set of experiments, electronic tongue data were compared to existing data from a human taste panel for several marketed products and prototype formulations. A good correlation (r(2)=0.99) was achieved from this comparison, and the relative taste of prototype formulations not tasted by humans was predicted.