Experimental work in animals and, to a more limited extent, in humans, has demonstrated that the cholinergic system is involved in mechanisms which control learning and memory. Since there is cholinergic loss in a variety of dementing illnesses, any treatment designed to alleviate the mental symptoms of these diseases must address the issue of cholinergic dysfunction even if other treatments are also required to overcome other neurotransmitter imbalances. Work in rodents has demonstrated that cholinergic-rich fetal neural tissue transplants can, under certain circumstances, alleviate the behavioral effects of cholinergic lesions or of cholinergic decline associated with aging. More complex cognitive testing can be achieved using primates and, in this case, the common marmoset is a suitable species to use because its rapid and reliable reproductive rate aids the provision of appropriate transplant tissue. Marmosets with transection of the fornix are deprived of a cholinergic input into the dentate gyrus, posterior hippocampus and entorhinal cortex and are specifically impaired on learning tasks which require remembering a rule of responding (non-evaluative memory). Transplantation of cholinergic-rich fetal septal tissue into the hippocampus of such animals completely restores their ability to learn this type of task, whereas transplantation of cholinergic-poor fetal hippocampal tissue into the same area produces no such improvements. These results demonstrate that where a learning impairment is produced by a relatively simple procedure which has a major effect on one neurotransmitter, that function can be restored by transplantation of tissue containing that neurotransmitter even where the impairment consists of a very "high level" cognitive dysfunction.