Split-brain monkeys were used to investigate the effects of inferior temporal lesions on visual discrimination behaviour. A lesion was made in only one hemisphere. Each hemisphere then behaved independently and similarly to either an animal with bilateral lesions or a normal animal, depending on which eye was open. Three types of removal were produced: a small neocortical lesion, a larger neocortical lesion and the latter combined with removal of the hippocampus. The 2 animals with small lesions were first taught a planimetric pattern discrimination then all animals were taught a series of common object discriminations. The animals were subsequently tested on the object discrimination after either 15 min or 20 h. In this way measures of the abilities of the animals to acquire a habit and to reproduce the behaviour after a short delay or a long delay were obtained. It was found that the small removal did not alter the rate at which the animals learned the objects discriminations but it did slow down learning on the pattern discrimination. The larger lesions significantly reduced the speed of learning on the object discriminations. However, it was found that all animals performed well when tested after either the short or the long delay. It was concluded that animals with temporal lobe damage have difficulty acquiring a memory-record underlying discrimination behaviour but having done so they are able to reproduce the behaviour after at least a 20-h interval. © 1969.