The effects of response topography on stimulus class formation were studied in two experiments. In Experiment 1, 32 college students were assigned to 2 response topographies and 2 stimulus sets, in a 2 x 2 design. Students selected stimuli by either moving a mouse to lace an arrow-shaped cursor on the stimulus or pressing a key corresponding to stimulus location. After they learned conditional discriminations EF, DE, CD, BC, and AB and a simple simultaneous discrimination, tests for class formation were conducted. The number of students showing class formation was larger for the group using the mouse. Stimulus set had no effect. Experiment 2 included probes for controlling relations in the baseline and showed that the response topography using the mouse increases the probability of sample-S+ relations, thus increasing the probability of class formation. (Contains 3 figures and 4 tables.)