The interaction between ethanol and caffeine on operant behavior was studied in 24 water-deprived male rats trained in a discrete trial spatial alternation schedule with water as reinforcer. One single drug dose-response experiment or one dose combination of ethanol and caffeine (including the associated control treatments) was run on 4 successive days in 1 week. The four treatments of 1 week were administered to each animal in a distinct order according to the 24 possible permutations. In the single drug weeks, ethanol (0.25, 0.5, 0.75 and 1.0 g/kg IP) or caffeine (5, 10, 20, and 40 mg/kg PO) were administered 15 min before the session. In four interaction experiments all combinations of two doses of ethanol (0.5 and 1.0 g/kg IP) and two doses of caffeine (25 and 50 mg/kg PO) were employed. Ethanol and caffeine alone showed both dose-dependently decreased choice accuracy and increased response latency and passiveness. In combination, caffeine normalized the ethanol-induced alterations in ITI response rate and pause length but potentiated the effects on choice accuracy, latency and number of pauses. The results are interpreted in terms of effects of these drugs on attentional and arousal processes, and the test procedure is proposed as a screening tool for the preclinical assessment of ethanol-drug interactions.