In a between-subject comparison of two memantine administration schedules we observed that treatment with the NMDA receptor antagonist memantine before testing sessions reduced ingestion of a 10% sucrose solution in rats, due to reduced licking burst size, thus suggesting a blunted hedonic response. Conversely, daily post-session administration reduced burst number, indicating a reduced level of behavioural activation, likely due to the development of conditioned taste aversion (CTA). In this study, the effect of pre-session and post-session memantine administration was investigated within-subjects. Memantine was administered in daily intraperitoneal injections for 13 days, on alternate days, either 1-h before-"before testing" sessions-or immediately after a 30-min session-"after testing" sessions. The effects on the microstructure of licking for a 10% sucrose solution were examined in the course of treatment and for 21 days after treatment discontinuation. The results show reduced burst size in the "before testing" sessions, without effects on the intra-burst lick rate, an index of motoric effects. Moreover, burst number was reduced since the third session of both administration conditions until the end of treatment. Interestingly, the effect of memantine of reducing the activation of ingestive behaviour was less pronounced in this study with respect to that observed with the previous study post-session administration schedule, in spite of the longer treatment. This apparent paradox might be explained if one considers these effects as instances of a memory-related effect, such as the development of CTA. In the framework of this hypothesis, the "before testing" sessions, not being followed by memantine administration, can be considered as extinction sessions performed every other day. Moreover, the animals treated with memantine at the highest dose failed to recover to pre-treatment ingestion levels 21 days after treatment discontinuation, while the animals treated after testing sessions in the previously published study showed a complete recovery well before the 15th day test. Within the same interpretative framework, this might depend by the reduced number and frequency of the extinction trials-i.e. the number of the sessions run after treatment discontinuation-in the present study. These results provide further support to the conclusion that memantine administration before sessions reduce burst size, an effect which is likely due to blockade of NMDA receptors occurring during behavioural testing. The observation that this effect can be obtained even in absence of a reduced intra-burst lick rate, which rules out the involvement of motor impairment, provides an important piece of evidence in support to the interpretation of this effect as a blunted hedonic response. Moreover, these results provide further evidence that burst number reduction is due to a memory-related effect induced by memantine administration after sessions.