Hippocampal long term potentiation (LTP) and depression (LTD) of glutamatergic transmission have been studied as cellular substrates for learning and memory. However, their exact roles in learning and memory are not well established, largely due to the lack of specific inhibitors for either LTP or LTD. It is generally accepted that the induction of both LTP and LTD at the CA1 synapse is postsynaptic and dependent upon Ca2+ influx through activated N-methyl-D-aspartate subtype glutamate receptors (NMDARs). However, the mechanisms underlying the expression of LTP and LTD remain hotly debated, and likely involve both a presynaptic component via alteration of transmitter release and a postsynaptic one through the modification of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid subtype glutamate receptors (AMPARs). Recent studies from many laboratories including our own have provided substantial evidence suggesting that AMPARs are continuously cycling between the plasma membrane and intracellular compartments via vesicle fusion mediated plasma membrane insertion and clathrin dependent endocytosis and that facilitated AMPAR insertion and endocytosis at postsynaptic membranes contributes to the expression of LTP and LTD, respectively. Using a combination of recombinant receptor expression systems and hippocampal brain slice preparations, we were able to demonstrate that facilitated endocytosis of postsynaptic AMPARs during LTD is AMPAR GluR2 subunit-specific (Man et al, 2000. These studies have lead us to develop a GluR2-derived interference peptide (GluR2-3Y) that, when delivered into neurons in the brain, can specifically block the expression of LTD without affecting the normal functioning of either NMDARs or AMPARs and hence basal synaptic transmission in many regions of the brain (Ahmadian et al, 2004). Importantly, we found that application of a membrane-permeant form of the GluR2 peptide (Tat-GluR2-3Y) specifically prevented certain LTD-dependent behaviours in rats in-vivo (Brebner et al, 2005). Availability of this systemic applicable, LTD-specific inhibitor allowed us for the first time to address specific roles of hippocampal CA1 LTD in hippocampus-dependent spatial learning and memory in the Morris water maze. We found that systemic administration of Tat-GluR23Y, but not a scrambled control peptide, 1h before the training phase of the task did not affect the rate of acquisition (learning) during the 8 training trials, but significantly impaired the hidden platform probing test (memory retrieval) performed on the second day. This impairment in memory retrieval was absent in the rat if peptide was given either immediately following training or before the probe test. Our findings strongly suggest that hippocampal LTD is induced during the learning phase or early phase of consolidation and that this LTD is necessary for the formation of long-term spatial memory. The present study also demonstrated the utility of peptides that disrupt AMPAR trafficking, the final step in the expression of synaptic plasticity, as tools to examine the critical role of LTD and/or LTP in specific aspects of learning and memory in conscious animals.