In recent years muscarinic receptor agonists and cholinesterase inhibitors have been developed for the treatment of Alzheimer's disease. We have evaluated examples from both classes of compounds in rodent tests of reference and working memory, as well as tests that are sensitive to the side-effects of these compounds. Thus, three selective muscarinic receptor partial agonists L-689,660, (M1/M3), AF102B (M1/M3) and L-687,306 (M1) and two cholinesterase inhibitors, E2020 and eptastigmine, were compared in a mouse tail-flick (TF) test, a rat response sensitivity (RS) test, in rat tests of reference memory, passive avoidance (PA) or conditioned suppression of drinking (CSD), and working memory (delayed-matching-to-position, DMTP). In the TF test, all of the compounds tested, with the exception of L-687,306, (1.0-30.0 mg/kg) dose-dependently induced antinociception of which L-689,660 was the most potent (minimum effective dose (MED) = 0.03 mg/kg). In the RS test, all of the compounds, but again with the exception of L-687,306, (1.0-30.0 mg/kg), dose-dependently reduced response rates, of which L-689,660 was again the most potent (MED = 0.1 mg/kg). In the reference memory test, all the compounds reversed the effects of a scopolamine-induced deficit with L-687,306 being the most potent (MED = 0.01 mg/kg). By contrast, in the DMTP test, although both the cholinesterase inhibitors and L-687,306 reversed the effects of scopolamine-induced deficit, L-689,660 and AF102B were without effects. These results suggest that cholinesterase inhibitors and low efficacy M1 selective muscarinic receptor agonists can reverse the effects of a scopolamine-induced deficit in animal tests of reference and working memory at doses that do not induce the side-effects usually associated with cholinomimetics.