The effects of disrupting the muscarinic or nicotinic systems on short-term spatial memory were investigated using a delayed matching to position (DMTP) procedure. Rats were trained on the DMTP until stability and then divided into two groups: one group was implanted with an indwelling cannula aimed at the lateral ventricle. The cannulated group received injections of selective muscarinic antagonists (pirenzepine, M1; AFDX 116, M2; UH-AH 37, M1/M3) or hemicholinium-3 (a choline uptake inhibitor). The remaining animals were treated with conventional muscarinic antagonists (scopolamine, methyl scopolamine) or nicotinic channel blockers (mecamylamine, hexamethonium). Scopolamine, methyl scopolamine and UH-AH 37 disrupted all performance parameters in a non-specific but dose related manner. Pirenzepine disrupted accuracy in a delay, but not dose dependent manner, and exerted no other negative effects on performance. Hemicholinium-3-induced performance deficits showed some elements of effects seen following pirenzepine and scopolamine (delay dependent effects on accuracy, some negative effects on other motoric aspects of performance). AFDX 116 and hexamethonium had no significant effects on performance with respect to control. Mecamylamine reduced accuracy and increased response latencies at the highest dose tested. These data indicate that muscarinic antagonists are more effective at disrupting mnemonic performance than nicotinic blockers, and moreover, that distinct muscarinic receptors may have differential effects on cognitive performance.