A series of rats received alcohol, caffeine, or alcohol-caffeine combinations. Conditioned and unconditioned responses as well as errors in choice (discrimination) were measured. No significant changes were noted in total conditioned or unconditioned responsiveness with small doses of caffeine (50 mg/kg) or with a dosage of 1 gAg of ethanol. By all measurements, as well as observation, caffeine potentiated the depressive effect of alcohol long after the alcohol had disappeared from the blood. This was most marked with intermediate (100 mgAg) or with high (150 mgAg) dosages of caffeine. In no instance did caffeine antagonize the effects of alcohol.