Inexperienced beer-drinkers who could not pass a beer-flavor discrimination test were divided into a taste group, which received additional perceptual experience with beer flavors; a verbal group, which received instruction in beer flavor terminology; a taste/verbal group, which received both additional experience and instruction; and a control group. Before and after training, all subjects participated in a similarity rating task involving beer flavors and flavor-related adjectives. Additional taste experience--but not increased experience with beer-flavor terminology--improved novices' ability to detect identical beer flavors. Results are discussed in terms of the trainability of flavor discrimination and the role of cognitive factors (both flavor-related and flavor-independent) in marketing beverages.