The current research examines the influence of midazolam (MDZ) on memory reconsolidation using a contextual fear paradigm in rats, based on three context-shock training trials (0.7 mA, 3 s). First, we evaluate the effect of MDZ (1 mg/kg, i.p.) injected shortly after the training procedure. Second, we examined the influence of MDZ after a brief exposure (90 s) either in the training context (reactivation procedure) or in a neutral environment (no reactivation procedure) and one day later, freezing behavior was scored when rats were re-exposed to the training environment. Third, we investigate both the effect of MDZ administered at different times following reactivation on fear memory and the persistence of such effect 10 days after reactivation. Finally, we test whether the MDZ effect could be reverted by a single weak training trial (0.2 mA, 3 s) or by the presentation of the same unconditioned stimulus in the absence of the conditioned stimulus as a reminder which proves to induce significant freezing in rats not previously trained. Results show that MDZ interferes with the formation of a contextual fear memory only when administered after the reactivation procedure but not after the training procedure. This interference was effective up to 60 min after reactivation and not at a later time. No spontaneous recovery of freezing behavior was observed 11 days after MDZ injection which was not reverted by a weak training trial and by the unconditioned stimulus alone. All these data support the idea that stimulating GABA A receptor sites via MDZ selectively disrupts the reconsolidation process of a contextual fear memory.