Two groups of rats (male, Long-Evans) were studied in a continuous access situation, in which ethanol, food and water intake patterns were monitored, 23 h/day. One group of rats was initiated to lever press for ethanol prior to study in the continuous access situation by use of a secondary-conditioning procedure. The other group had no prior initiation. It was found that the ethanol self-administration pattern of the initiated rats was similar to a previous study using another initiation procedure. After four weeks, the noninitiated group also demonstrated an ethanol intake pattern similar to initiated animals. However, the specific nature of individual ethanol drinking bouts in the noninitiated animals was found to be different, suggesting that initiation resulted in larger and more ethanol-drinking bouts. In addition, the noninitiated animals failed to show any home-cage shift in ethanol preference, which is observed after the use of the initiation procedures. The only major difference found between the sucrose-substitution initiation procedure and the secondary-conditioning procedure occurred when response requirements to obtain food were increased. In this situation, ethanol intake increased only in the sucrose-substitution initiated animals. The relation of this finding to the underlying theoretical basis for each type of initiation procedure is discussed.