This study extends the pharmacological characterization of a novel quantitative murine behavioral method, the mirrored chamber aversion assay, which appears to be selectively sensitive to anxiolytic agents. Behavioral effects of acute diazepam administration were compared with those of the 5-HT1A anxiolytic buspirone and those of ethanol in C57BL/6J mice. These known anxiolytics produced a dose-dependent reduction in avoidance behavior of large magnitude, as evidenced by statistically significant decreases in latency to enter and increases in time spent in the mirrored chamber. Anxiolytic-like effects of these compounds in the mirrored chamber assay differed from those observed by the elevated plus-maze method. The behavioral effects of the test compounds were not due to alteration of locomotor activity. These findings indicate that the murine mirrored chamber assay responds to several agents known to be anxiolytic in man but differs from the plus-maze in the pharmacological spectrum of the anxiolytics to which it is sensitive.