Although caffeine may be the most widely used behaviorally active drug, few studies have examined its rewarding properties. In the present study, the designs of place-conditioning and taste-conditioning paradigms were combined in a single experiment to provide two independent measures of drug reward. During 3 preconditioning sessions, undrugged rats received access to 2 distinctive chambers connected by a small tunnel. During the 8-session conditioning phase, groups were given home cage access to either a sweet or salty solution, administered caffeine (0.3, 1.0, 3.0, 10.0, 30.0 mg/kg IP; 30.0 mg/kg SC), and confined to one of the chambers. On alternate sessions, rats were given home cage access to the remaining solution, injected with the vehicle, and confined to the opposite chamber. On test sessions 1 and 2, undrugged animals were given home cage access to one of the flavored solutions and water, and then allowed access to both conditioning chambers. On test session 3, rats were given access to both flavored solutions, injected with the drug used during conditioning, and again allowed access to both chambers. Caffeine (3.0 mg/kg) produced a significant place preference. The highest dose (30.0 mg/kg IP and SC) produced significant place and taste aversions. A control group given (+)-amphetamine illustrated a significant place preference and taste aversion as expected. Thus caffeine appeared to produce a dose-dependent biphasic effect; a lower dose was rewarding, whereas higher doses produced aversions to environmental stimuli associated with the drug.