Use of the selective immunotoxin; 192 IgG-saporin, is helping to elucidate the role of the cholinergic system in cognition by overcoming the problems of interpretation associated with the use of non-specific lesioning agents. In separate studies, we have compared the long- and short-term effects of single site and combined saporin lesions of the nucleus basalis magnocellularis and medial septal area, on spatial learning and memory in radial arm and water maze tasks. At 11 months, only rats with combined lesions showed deficits in both radial and water maze tasks, although terminal cholinergic deafferentation was substantial and extensive tissue loss was seen at the injection sites in both single and combined lesions. However, the extensive tissue loss with long-term lesions suggested that behavioural deficits were not solely attributable to cholinergic deafferentation. In contrast, when rats with combined lesions were tested 5 months after lesioning, no deficits were apparent, although there was almost complete loss of choline acetyltransferase- and nerve growth factor receptor-immunoreactivity in the basal forebrain with no tissue damage at the injection sites. This study supports existing literature that selective loss of cholinergic neurons in the basal forebrain does not produce behavioural impairments in standard tasks of learning and memory, but deficits are apparent when damage is non-selective as occurs late after lesioning, confounding interpretation of behavioural data. It further highlights potential problems with this immunotoxin in long-term studies.