Spiperone reduced latency to feed and prolonged the total duration of feeding in a food-preference test in which novel and familiar foods were presented simultaneously to food-deprived rats. The increase in feeding duration could be accounted for in terms of an increase in time devoted to eating the novel foods, leaving the time devoted to eating the familiar food relatively unaltered. However, closer analysis of the data revealed that, at the level of mean duration of individual eating episodes, spiperone prolonged the eating-episode duration equally for both novel and familiar foods. This apparent paradox was resolved by noting that whilst spiperone did not alter the number of eating episodes in the test, it did alter the ratio of eating episodes in favour of novel foods. The data was incorporated into a control system model which depicts actions of spiperone in the food-preference situation with total eating duration as the final behavioural output.