To clarify which sites of the central nervous system are influenced by acute administration of alcohol at an early stage, auditory event-related potentials (ERP) using a target-selection paradigm were measured in 13 healthy volunteers. In the recorded waveforms of ERPs, the N100, P165, N200 and P300 latencies for the target tone and the N100 and P200 latencies for the non-target tone were determined in each subject before and 1- and 2-hr after ingestion of 200 ml of alcohol (containing 25% ethanol) or of 200 ml of water, for a total of six times on two different days. The P300 latency was significantly prolonged at 1- and 2-hr after alcohol ingestion; and, the N200 latency was significantly prolonged at 2-hr after alcohol ingestion. The 2-hr alteration in the P300 latency after alcohol ingestion was positively correlated with the ethanol dose per body weight. These data suggest that ethanol, in proportion to its dose, affects cognitive function estimated by the P300 latency earlier than other lower central nervous system functions. Evaluating subclinical effects on central nervous function, using the ERPs, of environmental neurotoxins such as organic solvents, researchers should pay particular attention not only to the degree of the drinking habit but also to the interval between the measurement and alcohol intake.