One theory holds that appetitive drives such as hunger and thirst are not conditionable because of their slow onset. However, recent evidence has shown only transitory conditioning of appetitive drives even with rapid onset. Such experiments may have failed because: (a) Exteroceptive conditioned stimuli (CSs) used in past experiments may be less easily accociated with the internal hunger drive than are interoceptive taste cues. Experiments 1-3 provided some support for this hypothesis. (b) The dependent measures used in past experiments may not be valid. Experiments 4 and 5 suggested that changes in the rate of bar pressing on an operant extinction curve following probe CSs for hunger may be a more sensitive and valid index of conditioned appetitive drive. However, the elusive and transitory nature of these results demands a reexamination of the basic difference between appetitive and aversive drives, which lies in the mode of their onset and control and which, given adaptive considerations, can account for their widely different conditionability.