In order to study whether paradoxical sleep (PS) is necessary to prepare subjects for the subsequent learning of a distributed two-way active avoidance conditioning, 10 rats were subjected to 5 h of paradoxical sleep deprivation (PSD group) by means of the platform method immediately before each of 5 acquisition sessions (one daily), as well as before a long-term retention (LTR) session (14 days). Another group of rats (PSD control group; n = 10) were placed on large platforms as a control for the side effects induced by PSD platforms. Rats in the dry control group (n = 10) did not receive any treatment. The number of avoidances of the PSD group was significantly lower on the 1st, 2nd and 3rd acquisition sessions compared to the PSD control group, and on the 2nd and 3rd sessions compared to the dry control group. PSD rats made significantly less intertrial crossings than dry controls on the 2nd and 3rd acquisition sessions, but no significant correlations were found between this variable and the number of avoidances. Therefore, our results are not fully in contradiction with the hypothesis that PS previous to the training sessions might prepare the animal for subsequent learning, although the influence of locomotor changes upon the performance of PSD subjects cannot be fully rejected.