A single exposure to the elevated plus-maze (EPM) test of anxiety reduces or abolishes the anxiolytic efficacy of benzodiazepines on a second trial. Some possible explanations to the occurrence of this phenomenon (one-trial tolerance-OTT) involve behavioral modifications thought to be consequence of some kind of learning in the first trial. In the present study, the influence of learning-impairing situations on the effects of the benzodiazepine chlordiazepoxide on mice re-tested in the EPM is investigated. The results showed that: (1) as expected, the administration of chlordiazepoxide to mice re-tested in the EPM- under the same conditions of the first trial- failed to induce anxiolysis; (2) a decreased percent time in the open arms was observed on the second trial of mice exposed to both trials under the same experimental conditions; (3) neither the increase in open arm avoidance by mice re-exposed to the EPM nor the OTT to chlordiazepoxide effect were modified by administration of the amnestic agent scopolamine; (4) the decrement of the duration of the first trial to 1 min or the change in light and noise conditions in both trials counteracted the increase in open arm avoidance on trial 2; (5) none of the later procedures modified the phenomenon of OTT. Although not discarding the modulation exerted by other memory processes in the OTT phenomenon, the results indicate that situations that impair the learned avoidance response to the open arms in the EPM do not modify the phenomenon of OTT.