1. Voltage-clamp experiments were used to study N-methyl-D-aspartic acid (NMDA) receptor antagonists applied by fast perfusion to mouse hippocampal neurons in dissociated culture. 2. Preincubation with the NMDA antagonists zinc (3-30 microM) and magnesium (30-300 microM) reduced subsequent responses to 100 microM NMDA applied together with these antagonists. No time dependence of antagonism was observed when responses were measured at the start and at the end of NMDA pulses 1.25-1.5 s in duration. 3. Two competitive antagonists of similar affinity in equilibrium experiments, D-2-amino-5-phosphonopentanoic acid (D-AP5) and 3-((+-)-2-carboxypiperazin-4-yl)propyl-1-phosphonic acid (CPP), had different profiles of action when applied as described above. With D-AP5, pulses of NMDA produced fast-on, fast-off responses, of reduced amplitude, similar to the effect of Zn and Mg. Responses to NMDA in the presence of CPP were also of reduced amplitude but, in addition, showed slow activation, such that the antagonist action of CPP decreased with time after the application of NMDA. 4. In the presence of 3 microM glycine and NMDA receptor antagonists with activity at the glycine modulatory site, either kynurenic acid (Kyn), 7-chlorokynurenic acid (7Cl-Kyn), or 5-chloro-indole-2-carboxylic acid (5Cl-I2CA), NMDA-evoked responses showed apparent use-dependent antagonism, such that the peak response to NMDA was much greater than the equilibrium response. A similar effect was produced by preincubation with low concentrations of glycine (less than 300 nM), which enhances desensitization of responses to NMDA. The apparent use-dependent action of glycine antagonists could be reversed on raising the glycine concentration and did not vary appreciably with changes in membrane potential over the range -60 to +50 mV. 5. Concentration-jump application of NMDA antagonists, in the presence of 100 microM NMDA and 3 microM glycine, were used to study antagonist association and dissociation kinetics directly. For D-AP5 and CPP, the dissociation rate was independent of antagonist concentration, and approximately 15 times faster for D-AP5 (19.6 s-1) than for CPP (1.36 s-1). The association rate for D-AP5 and CPP increased with antagonist concentration in a linear manner over the range 3-30 microM and was slower for CPP than for D-AP5, consistent with their similar potency at equilibrium. 6. In contrast to results obtained with CPP and D-AP5, the association rate for 7Cl-Kyn was approximately 3 times slower than the dissociation rate and did not change with concentration of antagonist.(ABSTRACT TRUNCATED AT 400 WORDS)