Subject performed a divided attention task in which they detected occu
rrences of a target letter in a stimulus figure which was a large lett
er made up of an array of small letters. Subjects pressed one button i
f the large letter was the target, another if the small letters were t
he target. Performance on this task was compared to other attention an
d psychomotor tasks in a three-way randomized crossover study comparin
g two doses of alcohol with placebo in 12 healthy volunteers (six male
, six female, aged 19-41). The higher dose of alcohol produced blood c
oncentrations of 43 mg/100 ml, the lower dose 16 mg/100 ml. Subjects r
eported themselves significantly more drunk on both doses of alcohol t
han on placebo, showed poorer long-term recall on the Buschke selectiv
e reminding task than on placebo, and made more errors on a letter can
cellation task. They performed letter cancellation faster, however. No
significant effects were seen on the divided attention or continuous
attention tasks, though continuous attention showed a trend towards im
pairment. These results suggest that the greater sensitivity showed by
some multiple task combinations to low doses of drugs such as alcohol
is not due to divided attention as such, but results from some aspect
of task combination. The coupling of increased errors but increased s
peed on letter cancellation has obvious implications for the interpret
ation of the effects of alcohol on complex real-life tasks such as dri
ving.