Septal neurons from embryonic rats were grown in tissue culture. Microfluorimetric and electrophysiological techniques were used to study Ca2+ homeostasis in these neurons. The estimated basal intracellular free ionized calcium concentration ([Ca2+]i) in the neurons was low (50-100 nM). Depolarization of the neurons with 50 mM K+ resulted in rapid elevation of [Ca2+]i to 500-1,000 nM showing recovery to baseline [Ca2+]i over several minutes. The increases in [Ca2+]i caused by K+ depolarization were completely abolished by the removal of extracellular Ca2+, and were reduced by approximately 80% by the 'L-type' Ca2+ channel blocker, nimodipine (1 microM). [Ca2+]i was also increased by the excitatory amino acid L-glutamate, quisqualate, AMPA and kainate. Responses to AMPA and kainate were blocked by CNQX and DNQX. In the absence of extracellular Mg2+, large fluctuations in [Ca2+]i were observed that were blocked by removal of extracellular Ca2+, by tetrodotoxin (TTX), or by antagonists of N-methyl D-aspartate (NMDA) such as 2-amino 5-phosphonovalerate (APV). In zero Mg2+ and TTX, NMDA caused dose-dependent increases in [Ca2+]i that were blocked by APV. Caffeine (10 mM) caused transient increases in [Ca2+]i in the absence of extracellular Ca2+, which were prevented by thapsigargin, suggesting the existence of caffeine-sensitive ATP-dependent intracellular Ca2+ stores. Thapsigargin (2 microM) had little effect on [Ca2+]i, or on the recovery from K+ depolarization. Removal of extracellular Na+ had little effect on basal [Ca2+]i or on responses to high K+, suggesting that Na+/Ca2+ exchange mechanisms do not play a significant role in the short-term control of [Ca2+]i in septal neurons. The mitochondrial uncoupler, CCCP, caused a slowly developing increase in basal [Ca2+]i; however, [Ca2+]i recovered as normal from high K+ stimulation in the presence of CCCP, which suggests that the mitochondria are not involved in the rapid buffering of moderate increases in [Ca2+]i. In simultaneous electrophysiological and microfluorimetric recordings, the increase in [Ca2+]i associated with action potential activity was measured. The amplitude of the [Ca2+]i increase induced by a train of action potentials increased with the duration of the train, and with the frequency of firing, over a range of frequencies between 5 and 200 Hz. Recovery of [Ca2+]i from the modest Ca2+ loads imposed on the neuron by action potential trains follows a simple exponential decay (tau = 3-5 s).