Cognitive impairments associated with schizophrenia await an effective treatment. In order to model schizophrenia-like cognitive deficits in rats, we evaluated the effects of ketamine, a dissociative anesthetic NMDA/glutamate receptor channel blocker in the attentional set-shifting task (ASST). Acute administration of ketamine (10 but not 3mg/kg) selectively impaired solving of the extradimensional (ED) set-shifting component. Next, we investigated whether the co-administration of mazindol, a dopamine and norepinephrine reuptake inhibitor would protect rats from ketamine-induced deficits. Mazindol dose-dependently and selectively alleviated ketamine-induced ED deficit with a minimal effective dose of 0.5mg/kg. The ED component improvement was noted primarily in ketamine - but not in vehicle co-treated rats, in which the drug facilitated ED shift solving at the dose as high as 5mg/kg. A "positive control", sertindole (2.5mg/kg) also ameliorated ketamine-induced ED deficit. Microdialysis of the prefrontal cortex in a separate group of animals revealed that 2-3h after the administration of 5mg/kg of mazindol and ketamine (i.e., at the time of ED component solving), the extracellular concentrations of dopamine were enhanced by ~300% as compared to the baseline and were intermediate between the mazindol- and ketamine-treated reference groups. However, at that time the levels of norepinephrine, serotonin and glutamate appeared unaffected. We conclude that ketamine may be useful in mimicking deficits specifically related to cognitive inflexibility observed in schizophrenia, and suggest that these anomalies could be ameliorated by mazindol. The beneficial effects of mazindol on ASST performance may have therapeutic implications for the treatment of schizophrenia.