2 experiments studied a 2-choice discrete RT task in which 2 unequally probable stimulus alternatives were assigned to 1 of a pair of unequally probable responses, and a single stimulus was assigned to the other. The relative frequencies of both stimuli and responses were systematically varied across conditions in Exp. I with 24 undergraduates. Exp. II with 8 Ss replicated the condition of greatest response frequency imbalance in Exp. I, but under instructions emphasizing accuracy. Comparisons of RTs to the 3 stimuli indicate that perceptual bias operates in this task and that response bias probably does not. (PsycINFO Database Record (c) 2006 APA, all rights reserved)