In order to investigate the effect of ethanol on visual attention, 18 subjects aged 20-50 years took part in a three-period crossover study in which they received placebo (PL) and two doses of ethanol in random order. The higher dose (E2: 0.88 g/kg, maximum 66 g for males, 55 g for females) was calculated to produce blood ethano concentrations of 60-80 mg/100 ml. The lower dose (E1) was 75 per cent of E2. Subjects showed highly significant subjective drunkenness at both doses (p < 0.01) and on the higher dose were slowed by 6-11 per cent on most speeded measures, in agreement with previous results. In a Four-Choice Reaction-Time Task, subjects responded for part of the time to a fixed, repetitive sequence, and at other times to a random sequence of stimuli. At the transition from repetitive to random sequences, subjects on ethanol showed a disproportionate slowing (60 per cent on the higher dose). This slowing may be of particular relevance to driving, as the time taken to engage control processing after a period of relatively automatic activity may be important in dealing with unexpected events on the road.