There is a critical need to develop novel pharmacotherapeutics capable of addressing the positive, negative, and cognitive symptoms of schizophrenia. Building on recent studies with a racemic mixture of the synthetic tetrahydroprotoberberine, D,L-Govadine, we isolated the D- and L-stereoisomers and employed a battery of behavioral, neurochemical, and electrophysiological procedures to assess their individual therapeutic potential. Rodent models predictive of antipsychotic efficacy and those that model positive symptoms were employed and we found that L-Govadine, but not D-Govadine, improved these measures. Pretreatment with either stereoisomer during CS pre-exposure prevented the disruption of latent inhibition by amphetamine. Moreover, pretreatment with either stereoisomer also improved deficits in social interaction in the neonatal ventral hippocampal lesioned rat. Improved cognitive performance in two different prefrontal cortex-dependent tasks was observed with D-, but not L-Govadine, which strongly suggests that the D-steroisomer may be an effective cognitive enhancer. Alterations in dopamine efflux were also assessed and L-Govadine increased dopamine efflux in both the prefrontal cortex and nucleus accumbens. However, D-Govadine only increased dopamine efflux in the prefrontal cortex and not in the nucleus accumbens. Electrophysiological studies confirmed that L-Govadine is a DA-D2 antagonist, whereas D-Govadine shows no appreciable physiological effects at this receptor. Collectively these data show that L-Govadine performs well on measures predictive of antipsychotic efficacy and rodent models of positive symptoms through antagonism of DA-D2 receptors, whereas D-Govadine improves impairments in compromised memory function in delayed response tasks possibly through selective increases in DA efflux in the frontal cortex.