Trained 8 groups of 12 female albino Sprague-Dawley rats to traverse a runway under 1 of 4 drive-reward combinations (deprived-food, deprived-no food, satiated food, or satiated-no food) and then shifted them to a different combination. During Phase I, only deprived-food groups showed increasing response speeds, and no energizing effect of drive was obtained for the deprived-no food group. During Phase II, the group which continued to show some consummatory activity even though it was food satiated extinguished more slowly then the groups from which food reward was omitted. For the other groups, satiated or nonrewarded training during Phase I had no effect upon performance during Phase II. Results of both phases support the view that consummatory activity is the necessary event in instrumental reward conditioning