A review is presented of experimental studies (using animal subjects) that investigated the effects of prior exposure to a pair of stimuli on subsequent visual discrimination learning. The bulk of these experiments used young rats as the subjects, with stimulus exposure taking place in the home cage over a long period of time. When such procedures produce transfer to later learning, it is usually positive. Also reviewed are related experiments that differ in their procedural details. (In some, the subjects are animals other than rats, the exposure occurs in environments other than the home cage, etc.) In most of these studies, later learning is found to be retarded. An attempt is made to isolate the critical feature or features of an exposure-learning experiment that determines whether transfer will be positive or negative, and possible theoretical interpretations of the empirical generalizations that emerge are also considered.