Research indicates opioid antagonists can reduce alcohol drinking in rodents. However, tests examining the effects of opioid antagonists on ethanol seeking and relapse behavior have been limited. The present study examined the effects of two opioid antagonists on ethanol maintenance, seeking, and relapse responding by alcohol-preferring (P) rats. Adult P rats were self-trained in two-lever operant chambers to self-administer 15% (vol/vol) ethanol on a fixed-ratio 5 (FR5) versus water on a FR1 concurrent schedule of reinforcement in daily 1-h sessions. After 10 weeks, rats underwent extinction training, followed by 2 weeks in their home cages. Rats were then returned to the operant chambers without ethanol or water to measure responses on the ethanol and water levers for four sessions. After a subsequent 2 weeks in the home cage, without access to ethanol, rats were returned to the operant chambers with ethanol and water available. Effects of antagonists on maintenance responding were tested after several weeks of daily 1-h sessions. Naltrexone (NAL; 1–10 mg/kg, subcutaneously [s.c.]; n = 8/dose), LY255582 (LY; 0.03–1 mg/kg, s.c.; n = 8/dose), or vehicle were injected 30 min before the first session (in the absence of ethanol), following 2 weeks in their home cages, and for four consecutive sessions of ethanol self-administration under maintenance and relapse conditions. Both NAL and LY reduced responses on the ethanol lever without any fluids present, and ethanol self-administration under relapse and on-going drinking conditions, with LY being more potent than NAL. Both NAL and LY were less effective in reducing responding in the absence of ethanol than in reducing ethanol self-administration. Overall, the results indicate that the opioid system is involved in mediating ethanol seeking, and ethanol self-administration under relapse and on-going alcohol drinking, but that different neurocircuits may underlie these behaviors.