With respect to the treatment of addiction, the objective of extinction training is to decrease drug-seeking behavior by repeatedly exposing the patient to cues in the absence of unconditioned reinforcement. Such exposure therapy typically takes place in a novel (clinical) environment. This is potentially problematic, as the effects of extinction training include a context dependent component and therefore diminished efficacy is expected upon the patient's return to former drug-seeking/taking environments. We have reported that treatment with the NMDAR coagonist D-serine is effective in facilitating the effects of extinction to reduce cocaine-primed reinstatement. The present study assesses D-serine's effectiveness in reducing drug-primed reinstatement under conditions in which extinction training occurs in a novel environment. After 22 days of cocaine self-administration (0.5 mg/kg) in context “A”, animals underwent 5 extinction training sessions in context “B”. Immediately after each extinction session in “B”, animals received either saline or D-serine (60 mg/kg) treatment. Our results indicate that D-serine treatment following extinction in “B” had no effect on either IV or IP cocaine-primed reinstatement conducted in “A”. These results stand in contrast to our previous findings where extinction occurred in “A”, indicating that D-serine's effectiveness in facilitating extinction training to reduce drug-primed reinstatement is not transferable to a novel extinction environment. This inability of D-serine treatment to reduce the context specificity of extinction training may explain the inconsistent effects observed in clinical studies published to date in which adjunctive cognitive enhancement treatment has been combined with behavioral therapy without significant benefit.