Electrophysiological and pharmacological methods were used to examine the role of glutamate in mediating the excitatory and inhibitory responses produced by the N2v rasp phase neurons on postsynaptic cells of the Lymnaea feeding network. The N2v → B3 motor neuron excitatory synaptic response could be mimicked by focal or bath application of L-glutamate at concentrations of ≤10-3 M. Quisqualate and α-amino-3-hydroxy-5-methyl-4- isoxazolepropionic acid (AMPA) were potent agonists for the B3 excitatory glutamate receptor (10-3 M), whereas kainate only produced very weak responses at the same concentration. This suggested that non-N-methyl-D- aspartate (NMDA), AMPA/quisqualate receptors were present on the B3 cell. The specific non-NMDA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10-5 M) blocked 85% of the excitatory effects on the B3 cell produced by focal application of glutamate (10-3 M), confirming the presence of non-NMDA receptors, CNQX also blocked the major part of the excitatory postsynaptic potentials on the B3 cell produced by spontaneous or current-evoked bursts of spikes in the N2v cell. As with focal application of glutamate, a small delayed component remained that was CNQX insensitive. This provided direct evidence that glutamate acting via receptors of the non- NMDA, AMPA/quisqualate type were responsible for mediating the main N2v → B3 cell excitatory response. NMDA at 10-2 M also excited the B3 cell, but the effects were much more variable in size and absent in one-third of the 25 B3 cells tested. NMDA effects on B3 cells were not enhanced by bath application of glycine at 10-4 M or reduction of Mg2+ concentration in the saline to zero, suggesting the absence of typical NMDA receptors. The variability of the B3 cell responses to NMDA suggested these receptors were unlikely to be the main receptor type involved with N2v → B3 excitation. Quisqualate and AMPA at 10-3 M also mimicked N2v inhibitory effects on the B7 and B8 feeding motor neurons and the modulatory slow oscillator (SO) interneuron, providing further evidence for the role of AMPA/quisqualate receptors. Similar effects were seen with glutamate at the same concentration. However, CNQX could not block either glutamate or N2v inhibitory postsynaptic responses on the B7, B8, or SO cells, suggesting a different glutamate receptor subtype for inhibitory responses compared with those responsible for N2v → B3 excitation. We conclude that glutamate is a strong candidate transmitter for the N2v cells and that AMPA/quisquate receptors of different subtypes are likely to be responsible for the excitatory and inhibitory postsynaptic responses.