Most previous studies investigating the relationship between N-methyl-D-aspartate receptor-dependent synaptic plasticity and learning have employed drugs that either compete with glutamate for access to the primary agonist binding site (e.g., D-2-amino-5-phosphopentanoic acid) or block the associated ion channel (e.g., dizocilpine). This study targeted the glycine receptor site located on the NMDA receptor complex. Chronic intracerebroventricular infusion of the glycine site antagonist 7-chlorokynurenate (7CK; 75 mM, 0.5 microliter/h, icv, for up to 14 days) impaired performance of male Lister hooded rats during acquisition of a spatial reference memory task in the water maze. In addition, however, these animals showed sensorimotor deficits, including a prolonged righting reflex, ataxia, and difficulty in staying on the escape platform. On completion of behavioral testing, the rats were anesthetized with urethane and an attempt was made to induce LTP in the hippocampus ipsilateral to the infusion cannula. Both control and 7CK-infused animals displayed equivalent long-term potentiation (LTP) 60 min posttetanus. A novel analytical technique for assaying drug tissue levels involving high-performance liquid chromotography with fluorescence detection revealed that tissue levels of 7CK in hippocampus were extremely low and unlikely to be sufficient to affect LTP, as observed. These findings neither support nor compromise the LTP/learning hypothesis, but they illustrate some of the problems of using drugs to elucidate the neurobiological mechanisms of learning and memory and the importance of a within-subjects design incorporating behavioral, physiological, and biochemical measures.