Group-housed female Sprague-Dawley rats were trained to self-administer 5% ethanol (v/v) in a large self-administration chamber (100 x 40 x 40 cm) following three different initiation methods. The procedures were 1) an ethanol injection procedure, 2) a sucrose substitution procedure, and 3) a prandial drinking technique. Only the prandial drinking method served to maintain responding for ethanol in the absence of water deprivation or sweetening of the alcohol solution. Rats trained using this technique showed a large preference for 5% ethanol over water and a significant increase in locomotor activity while responding for 5% ethanol but not while responding for water. When the concentration of ethanol was increased from 1% to 32%, the amount of ethanol ingested increased up to a maximum of 1.233 +/- 0.3 g/kg of 32% ethanol, and response rates and number of ethanol deliveries followed an inverted U-shaped curve. Appreciable blood ethanol levels were detected immediately following self-administration of 8% ethanol. These results show that, in female Sprague-Dawley rats under the experimental conditions described, the prandial drinking technique was the most effective in inducing stable oral ethanol self-administration and suggest that under these conditions and in these subjects ethanol was acting as a positive reinforcer.