The Multiple Sleep Latency Test (MSLT) was used to assess the effects of ethanol at the peak and descending phases of the breath ethanol curve. Ethanol (0.75 g/kg) was administered (at 0900 hr) to 8 healthy, normal-sleeping men, aged 21 to 45 years old after 8 hr of sleep the previous night. MSLTs were conducted and breath ethanol concentrations (BrECs) were measured at 15, 45, 75, 105, 225, and 345 min after drinking was completed. Subjective measures were administered immediately before each sleep latency test. BrECs over the first 75 min (tests 1 to 3) peaked and differed from all subsequent tests (tests 4 to 6) over which BrECs declined. Sleep latency and subjective measures were averaged over tests 1 to 3 and 4 to 6. There was a significant increase in mean sleep latency relative to placebo for tests 1 to 3 and a significant reduction for tests 4 to 6. The subjective measure of stimulation sedation, the Biphasic Alcohol Effects Scale, showed lessened sedation after ethanol versus placebo on tests 1 to 3, compared with tests 4 to 6. This study confirmed the presence of a biphasic ethanol effect using an electrophysiological method (MSLT), showing increased physiological alertness on the peak phase of the BrEC curve and increased sedation on the descending phase. Relative to the effects observed on the MSLT with other low-dose stimulant drugs, the stimulatory effect of ethanol was mild.