Twelve rats were trained to press one lever after cocaine injection (3 mg/kg i.p.) and another lever after saline injection. Once rats were reliably discriminating cocaine from saline, other drugs were examined for their efficacies in substituting for cocaine. The dopamine uptake inhibitors WIN 35,428 [2-beta-carbomethoxy-3-beta-(4-fluorophenyl)tropane-1,5-naphthalene - disulfonate] and GBR 12909 (1-[2-bis(4-fluorophenyl)methoxy]ethyl]-4-[3- phenylpropyl]piperazine dihydrochloride) fully substituted for cocaine (cocaine responding > 80%), whereas the peripherally active cocaine methiodide and the 5-hydroxytryptamine uptake inhibitor fluoxetine did not substitute at all. Pentobarbital also failed to produce any cocaine-appropriate responding. Two selective norepinephrine uptake inhibitors were tested: tomoxetine fully substituted for the 3-mg/kg dose of cocaine and nisoxetine approached full substitution (79.7% cocaine responding). The direct-acting dopamine D-1 agonists SKF 38393 [(+-)-7-bromo-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-be nzazepin e HCl], SKF 77434 [(+-)-7,8-dihydroxy-3-allyl-1-phenyl-2,3,4,5-tetrahydro-1H-benzazepine HCl] and SKF 75670 [3-methyl-7,8-dihydroxyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benza zep ine HBr] fully substituted for cocaine, whereas the peripherally active dopamine D-1 agonist fenoldopam did not. Of four dopamine D-2 agonists tested, only quinpirole fully substituted; the others (N-0434 [(+-)-2-(N-propyl-N-phenylethylamino)-5-hydroxytetralin], (-)-NPA [R(-)-propylnorapomorphine HCl] and SDZ 208-912 (N-[(8-)-2,6-dimethylergoline-8-yl]-2,2-dimethyl-propanamide)) produced very limited partial substitution (cocaine responding < 32%).(ABSTRACT TRUNCATED AT 250 WORDS)