Tolerance to ethanol was tested with 24 adult, male, Wistar rats using a modified Geller-Seifter paradigm. This design consisted of three components: an unpunished random interval-30 sec (RI-30) schedule, a time-out period, and a punished plus continuous reinforcement (conflict) schedule. This procedure allowed for a distinction to be made between the sedative and the "anxiolytic" effects of ethanol. Rats were trained on this procedure until response rates stabilized. They were then randomly divided into three groups. One group served as a control group for two experiments. In the first experiment, the test group received ethanol (0.75 g/kg IP) on days 2-8 and saline on days 1 and 9. For the second experiment, the control group received saline and the test group received ethanol three times in one day at three hour intervals. The "anxiolytic" action (defined as an increase in the rate of punished responding or antipunishment effect) was observed to undergo rapid tolerance (by day 4 in experiment one and by the second session in experiment two), while tolerance to the sedative action (defined as a decrease in the rate of unpunished responding) developed more slowly (only by day 7 in experiment one, but by the third session in experiment two). These results suggest that rapid tolerance develops to the "anxiolytic" actions of ethanol and this may have important implications for the development of ethanol abuse.