Trained 7 stumptail and 4 rhesus monkeys on a successive black-white discrimination (e.g., black go left, white go right) and then gave them transfer tests to intermediate gray stimuli. Both response latencies and Ss' judgments were recorded. Consistent with other studies and a "no vicarious trial and error" (NO-VTE) model of choice behavior, Ss' judgments of the gray stimuli were graded rather than dichotomous. Contrary to the NO-VTE model, response latencies were curvilinearly related to the shade of gray. A simple "vicarious trial and error" model accounted for both main features of the data. The latter model also suggested that S's initial left-right orientation was an important factor in determining whether S's judgments of the gray stimuli were biased toward black or white. Response latencies of individual Ss support this implication. (PsycINFO Database Record (c) 2006 APA, all rights reserved)