In order to test the theory that nicotine is a reinforcing constituent in tobacco, smokers' nicotine preference was assessed after two hours' cigarette deprivation and immediately after smoking two cigarettes. Preference was measured by allowing subjects to freely adjust the nicotine concentration of each puff using a smoke mixing device. Nicotine preference was significantly higher after deprivation, showing that nicotine in cigarette smoke is positively reinforcing and smokers' attempts to obtain nicotine vary with prior cigarette consumption.