Ethanol's post-training facilitation of memory was examined using a latent learning paradigm known as the "water-finding task." Rats were assigned to one of two ethanol groups (E0.75 g/kg or E1.5 g/kg) or to a control group (saline) and individually placed in a novel open field containing a drinking tube. Following this exposure, subjects were immediately administered intraperitoneal (IP) injections of either the saline or ethanol and 48 hours later, re-introduced to the field. Initial latencies to contact the tube each time were recorded. A linear regression analysis of trial 2 latencies regressed onto trial 1 latencies indicated a statistically significant effect of ethanol on the relation between initial and subsequent latencies. Though the control rats' trial 2 latencies were completely random with respect to their previous speeds (rSAL = -0.07), the ethanol rats' trial 2 latencies were positively correlated with initial speeds (rE0.75 = 0.35, rE1.5 = 0.67). These results suggest that under conditions of post-training ethanol, trial 2 behavior is more similar to, or controlled by, trial 1 behavior and are consistent with the argument that, under certain training and testing contexts, ethanol can come to exert control over a response's recurrence.