Objective It has been suggested that the anterior cingulate cortex (ACC) and prelimbic cortex (PL) of medial prefrontal cortex (mPFC) have critical roles in the acquisition and retrieval (expression) processes of trace eyeblink conditioning (TEC), respectively.It was also reported that re-training after PL lesions following initial leaming could recover the impaired behavioral retrieval.Based on these results, we hypothesized that ACC substituted for the impaired retrieval role of PL in TEC.The present study is to investigate (1) the hypothesis of TEC retrieval substitution and (2) the natural roles of ACC and PL in TEC.Methods Lidocaine hydrochloride (LH) or saline was microinfused into ACC, or PL, or both ACC and PL after daily pre-infusion session training following 8 days initial acquisition training.Results (1) Infusion of lidocaine into area 24 does not significantly affect TEC expression.(2) Area 24 infusions affect TEC acquisition as evidenced by a difference in pre-infusion and post-infusion performance between sham and lesion groups that develops over several infusion days.(3) Area 32 infusions have immediate effect on TEC expression.(4) The expression deficits with area 32 infusions fades over 5 infusion sessions, and (5) subsequent dual infusions into both regions produces immediate deficits that do not recover.Conclusion The results support our hypothesis of substitution and also indicate that reversible inactivation of ACC prevented the acquisition of TEC but did not affect its retrieval, whereas reversible inactivation of PL caused significant deficits in retrieval of TEC, but did not affect its acquisition.