Forty male subjects were divided into two groups on the basis of their scores on a computerized intelligence test administered in a preliminary session. They then participated in a text recall and recognition experiment in which they received alcohol (1.0 ml/kg body weight) in one session and no alcohol in another session (the order being counterbalanced). Subjects were required to read a short passage, immediately recall as much of it as possible and then select from sets of four sentences the ones that appeared in the original passage (recognition). Although two different passages were used, and subjects knew about the memory tasks in advance, there was nevertheless a significant practice effect: subjects recalled 10.6 per cent more propositions on the second occasion than on the first, but this could be at least partly explained by the fact that they spent 10.0 per cent longer reading the second passage. There was also a significant 7.8 per cent slowing of reading time due to alcohol. Practice did not interact with either intelligence or alcohol on any measure of performance. Alcohol impaired the performance of high intelligence test scorers more than low scorers. The effects of intelligence and alcohol on free recall were quantitatively similar: high intelligence test scorers recalled 8.7 more propositions than low scorers; with alcohol, subjects recalled 8.3 fewer propositions than without alcohol. However, the effects were qualitatively different. The effect of intelligence was primarily on the recall of lower level propositions, whereas alcohol impaired the recall of both higher and lower level propositions. Furthermore, while there was no effect of intelligence on recognition, alcohol impaired performance such that more incorrect sentences were erroneously recognized, particularly those which altered the meaning of the original sentence, rather than merely its surface structure. In terms of Kintsch & van Dijk's (1978) model of text processing, the results are interpreted as suggesting that, while both intelligence and alcohol affect the capacity of the short-term memory buffer, there is an additional impairment with alcohol in the ability to select between higher and lower level propositions prior to filling the buffer.