elicited increased alcohol self-administration in both groups of alcohol-dependent animals compared to baseline alcohol self-administration and compared to nondependent animals (post-vapor behavioral data was averaged from self-administration sessions over 5 days before day 77 Set I or Set II). Specifically, alcohol intake (g/kg) was increased post-vapor in dependent animals compared to alcohol intake (g/kg) post-air control in nondependent animals (Set I: F3,20 = 5.4, p = 0.008; Fig. 1b; Set II: F3,19 = 5.9, p = 0.006; Fig. 1d). Furthermore, dependent animals had increased lever responding for alcohol post-vapor compared to pre-vapor (baseline) and compared to lever responding for alcohol in nondependent animals post-air control (post-vapor lever presses for alcohol, average day 70-75, Set I: F3,15 = 5.6, p = 0.01; Fig. 1c; Set II: F3,19 = 11.4, p = 0.0003; Fig. 1e). Alcohol vapor chambers were set to elicit blood alcohol levels averaging 150-200 mg% in the dependent groups. As expected, nondependent groups (exposed only to air) had no reliable blood alcohol levels: (dependent: (Set I: 193.2 ± 17.7 mg%; Set II: 159.7 ± 9.7 mg%; p = 0.14) versus nondependent: Set I and II combined, 10 ± 0.5 mg%, F = 227.6, p < 0.0001). Previous work shows
