The rate of Mn(2+)-induced fluorescence quenching (RFQ) was used as a relative measure of plasma membrane Ca2+ permeability (PCa) in fura-2-loaded cultured hippocampal neurons and cerebellar granule cells during and after protracted (15-30 min) glutamate (GLU) treatment. Some limitations of this method were evaluated using a kinetic model of a competitive binding of Mn2+ and Ca2+ to fura-2 in the cell. In parallel experiment a contribution of Ca2+ influx to the cytoplasmic Ca2+ ([Ca2+]i) was repeatedly examined during and following a prolonged GLU challenge by short-duration "low-Ca2+ trials" (50 microM EGTA) and by measurements of 45Ca2+ uptake. Experiments failed to reveal a putative persistent increase in PCa that earlier was thought to underlie Ca2+ overload of the neuron caused by its toxic GLU treatment. By contrast, a sustained increase of [Ca2+]i was found to be associated with a progressive decrease in PCa and Ca2+ influx both in the period of GLU application and after its termination. These findings give new evidence in favour of the hypothesis that the GLU-induced Ca2+ overload of the neuron mainly from an impairment of its Ca2+ extrusion systems.