Coffee-drinking cigarette smokers take in more nicotine when they ingest almost no caffeine than when they ingest an amount of caffeine ranging from 75 mg to 300 mg. They do not take in relatively less nicotine as the dose of caffeine increases from 75 mg to 300 mg. It seems, then, that something due to caffeine deficit is responsible for the effect. Heavier users of caffeine show this effect less strongly than do lighter users of caffeine. These results are discussed in terms of the discriminability of caffeine and nicotine deficits and the possible influence of differential tolerance to caffeine. The importance of evaluating caffeine consumption when studying nicotine use and the importance of considering the chronic level and use of these drugs when studying their effects on behavior is indicated.