The rats were self-trained in 3-lever operant chambers using essentially the same procedure previously described for a 2-lever operant paradigm (Rodd et al, 2008; Rodd-Henricks et al., 2002a, b). The P rats (n = 7) were given 4 1-hour sessions in the 3-lever operant chambers to respond for 10, 20, and 30% ethanol on an FR5 schedule of reinforcement. Water was freely available during these sessions. Water and food were freely available in the home cage. The 1-hour operant sessions were conducted during the dark cycle; there was a 1-hour interval between each operant session. Sessions were conducted 5 consecutive days each week. Following 10 weeks of ethanol access, rats were taken through 2 cycles of 2 weeks of alcohol deprivation and 2 weeks of ethanol access. Ethanol intakes were similar during the relapse periods as were observed during weeks 9 and 10. Rats were killed by decapitation 3 hours after the 4th session at the end of the second 2-week relapse period. This 3-hour time point was selected in an attempt to maximize the response to alcohol on the
