Rats were trained in one-trial step-down inhibitory avoidance and tested either 3 h or 31 days later. Ten minutes prior to the retention test, through indwelling cannulae placed in the CA1 region of the dorsal hippocampus, they received 0.5 microl infusions of: saline, a vehicle (2% dimethylsulfoxide in saline), the glutamate NMDA receptor blocker, aminophosphonopentanoic acid (AP5) (5.0 microg), the AMPA/kainate receptor blocker, cyanonitroquinoxaline dione (CNQX) (0.25 or 1.25 microg), the metabotropic receptor antagonist, methylcarboxyphenylglycine (MCPG) (0.5 or 2.5 microg), the inhibitor of calcium/calmodulin-dependent protein kinase II (KN62) (3.5 microg), the inhibitor of cAMP-dependent protein kinase (PKA), Rp-cAMPs (0.1 or 0.5 microg), the stimulant of the same enzyme, Sp-cAMPs (0.1 or 0.5 microg), or the inhibitor of the mitogen-activated protein kinase (MAPK) kinase, PD098059 (10 or 50 microM). CNQX, KN62 and PD098059 were dissolved in the vehicle; the other drugs were dissolved in saline. All these drugs, at the same doses, had been previously found to affect short- and long-term memory formation of this task. Retrieval measured 3 h after training (short-term memory) was blocked by CNQX and MCPG, and was unaffected by all the other drugs. In contrast, retrieval measured at 31 days was blocked by MCPG, Rp-cAMPs and PD098059, enhanced by Sp-cAMPs, and unaffected by CNQX, AP5 or KN62. The results indicate that, in CA1, glutamate metabotropic receptors are necessary for the retrieval of both short- and long-term memory; AMPA/kainate receptors are necessary for short-term but not long-term memory retrieval, and NMDA receptors are uninvolved in retrieval. Both the PKA and MAPK signalling pathways are required for the retrieval of long-term but not short-term memory.