Rats were trained to discriminate physostigmine (0.1 mg/kg s.c.) from saline in a two-choice, discrete-trial avoidance paradigm. Stimulus generalization curves for physostigmine were steep; complete generalization with physostigmine occurred only at the 0.1 mg/kg training dose. The muscarinic cholinergic agonists oxotremorine, pilocarpine and arecoline were evaluated for physostigmine-like discriminative effects. Oxotremorine generalized with physostigmine at a dose of 0.1 mg/kg in four of six rats tested; partial generalization was engendered by this dose in the remaining two animals. Complete generalization with physostigmine was produced in only one of six rats treated with pilocarpine. However, pilocarpine (3.0-10 mg/kg) did engender some physostigmine-appropriate responding in all rats tested. Arecoline (1.0 mg/kg) produced primarily saline-appropriate responding in all animals tested. Neostigmine, eseroline and nicotine were tested for physostigmine-like discriminative effects in order to assess the specificity of the physostigmine cue. Eseroline (1.0 mg/kg), an opioid-like derivative of physostigmine, and neostigmine (0.1 mg/kg) engendered saline-appropriate responding. Similarly, nicotine failed to generalize with physostigmine at doses up to 1.0 mg/kg. The discriminative stimulus effects of physostigmine were sensitive to antagonism by atropine. Complete blockade of the stimulus effects of the training dose of physostigmine was produced by 1.0 to 3.0 mg/kg of atropine. In contrast, a 10-fold higher dose of the quaternary antagonist homatropine methylbromide was necessary to block the discriminative effects of physostigmine. The discriminative effects of physostigmine were not blocked or were only partially blocked by mecamylamine at doses up to 10 mg/kg. The results of these experiments suggest that the discriminative stimulus effects of physostigmine are selective and probably centrally mediated.(ABSTRACT TRUNCATED AT 250 WORDS)