This paper reviews five separate experimental studies concerning the acute or chronic effects of alcohol on late ERP components related to cognitive processes. The studies are evaluated for their adequacy in two arenas: (a) task relevance from a behavioral viewpoint, and (b) their ability to clarify the functional significance of alcohol-related ERP changes. Based upon the general trends of the results reported from the five individual laboratories conducting the investigations, it can be concluded that the P3 component is the most consistently altered of the four late ERP components considered (N1, N2, P2, P3). The finding of a reduced P3 amplitude was more reliable than the finding of increased P3 latency. Alterations in ERP components, when they did occur, occurred under the acute influence of ethanol, as well as in abstinent chronic alcoholics. All of the ERP findings reported were obtained using variations of signal detection procedures. Changes in ERP components did not always parallel performance changes, suggesting that aspects of psychological processes (e.g., attention and stimulus evaluation), presumed to underly ERPs' functional significance, did not directly reflect these processes.