The anatomical substrate for coordinated release from the dispersed gonadotropin-releasing hormone (GnRH) neuronal population remains obscure. There is physiological evidence that the GnRH hormone itself has a role in tonic inhibition or modulation of GnRH function. This has led to the hypothesis that there is an ultrashort negative feedback mechanism subserved by axon collaterals acting back on the GnRH neurons. Recent ultrastructural studies have revealed GnRH synapses on GnRH neurons and their processes. Furthermore, there are alterations in the frequency of these synapses with the age and hormonal condition of the animal. Another candidate for coordination of neuronal activity for which there is some evidence in the magnocellular system, is the gap junction. Recently, physiological and anatomical evidence for gap junctional modifications among an immortalized GnRH-secreting cell line (GT1) has been reported. However, at present there is no immunocytochemical or ultrastructural evidence for gap junctions between GnRH neurons. A third and highly unorthodox anatomical relationship between (among) these cells has been suggested by serial ultrastructural reconstructions of pairs of GnRH neurons in close association. In some regions, GnRH neuronal processes can be seen to extend from each member of a pair of GnRH neurons. These meet and merge, forming an intercellular bridge. This phenomenon has been observed in several pairs of GnRH neurons in rat and monkey. The important caveat in making these observations is that techniques employed to demonstrate sites of antigenicity can severely compromise the ultrastructural integrity of membrane components. For this reason, further verification of the existence of intercellular bridges is being pursued. Should their existence be confirmed, they would be prime candidates for the coordination of secretory events among the scattered GnRH neuronal population.