1. The dynamic and kinetic interactions of alcohol and caffeine were studied in a double-blind, placebo controlled, cross-over trial. Treatments were administered to eight healthy subjects in four experimental sessions, leaving a 1 week wash-out period between each, as follows: 1) placebo, 2) alcohol (0.8 g kg-1), 3) caffeine (400 mg) and 4) alcohol (0.8 g kg-1) + caffeine (400 mg). 2. Evaluations were performed by means of: 1) objective measures: a) psychomotor performance (critical flicker fusion frequency, simple reaction time and tapping test), b) long latency visual evoked potentials ('pattern reversal'); 2) subjective self-rated scales (visual analogue scales and profile of mood states); 3) caffeine and alcohol plasma concentration determinations. 3. The battery of pharmacodynamic tests was conducted at baseline and at +0.5 h, +1.5 h, +2.5 h, +4 h and +6 h. An analysis of variance was applied to the results, accepting a P < 0.05 as significant. The plasma-time curves for caffeine and alcohol were analysed by means of model-independent methods. 4. Results obtained with caffeine in the objective measures demonstrated a decrease in simple reaction time and an increase in the amplitude of the evoked potentials; the subjects' self-ratings showed a tendency to be more active. Alcohol increased simple reaction time and decreased amplitude of the evoked potentials, although the subjects rated themselves as being active. The combination of alcohol + caffeine showed no significant difference from placebo in the objective tests; nevertheless, the subjective feeling of drunkenness remained. The area under the curve (AUC) for caffeine was significantly higher when administered with alcohol.(ABSTRACT TRUNCATED AT 250 WORDS)