Unlike other CNS depressants, alcohol intoxication can be associated with increased error rates, coupled with unaffected (or speeded) response rates during psychomotor and cognitive processing. The present study examined whether concurrent consumption of caffeine may differentially affect these aspects of alcohol and performance.A randomised, double-blind, placebo-controlled design was utilised in which 64 healthy young volunteers received either 0.66 g/kg alcohol, caffeine (110-120 mg), both or neither. Performance was assessed using a four choice reaction time task (FCRT) with elements of repetitive (predictable) and random stimuli sequences and the digit symbol substitution task (DSST). Individuals on alcohol made significantly more errors during both fixed and random FCRT sequences, and there was evidence of weak antagonism of these effects by caffeine on the latter measure. On the DSST test of psychomotor speed, alcohol was associated with a significant slowing, the caffeine group were significantly faster and there was clear antagonism of the effects of alcohol by caffeine. These findings confirm that alcohol consumption is associated a greater number of errors and provide some evidence for task-specific antagonism of alcohol's cognitive effects by caffeine.