Aging-, disease- and medication-related imbalance of central dopaminergic neurons causes functional impairment of cognition and neuropsychological delirium in humans. We attempted to develop a new delirium model using the direct dopamine agonist, apomorphine, and a choice reaction performance task performed by middle-aged rats. The psychological properties of the model were assessed by determining behavioral measures such as choice reaction time, % correct and % omission. Apomorphine (0.03-0.3 mg/kg s.c.) produced a dose-dependent impairment of task performance. The dose of 0.1 mg/kg prolonged choice reaction time, decreased % correct and increased % omission, indicating that rats had attentional deficits and a reduced arousal or vigilance but no motor deficits or reduced food motivation. This psychological and behavioral impairment of performance resembled that of clinically defined delirium. In this model, the cholinomimetic, aniracetam (10 mg/kg p.o.), reversed the performance impairment induced by apomorphine. Its two metabolites, 2-pyrrolidinone (10 and 30 mg/kg p.o.) and N-anisoyl-gamma-aminobutyric acid (GABA, 10 mg/kg p.o.), effectively reversed the performance impairment as the intact drug did. Another pyrrolidinone derivative, nefiracetam (10 and 30 mg/kg p.o.), tended to worsen the apomorphine effect. The cholinesterase inhibitor, tacrine (10 mg/kg p.o.), markedly worsened all of the behavioral measures. Neuroleptics, haloperidol (0.025 mg/kg s.c.), tiapride (30 mg/kg p.o.) and sulpiride (10 and 30 mg/kg p.o.), antagonized the apomorphine effect. The present results suggest that apomorphine-induced behavioral disturbances in the choice reaction performance task seems to be a useful delirium model and aniracetam may improve delirium through the action of 2-pyrrolidinone and N-anisoyl-GABA, presumably by facilitating dopamine release in the striatum by acting as an AMPA or metabotropic glutamate receptor agonist.