We investigated the mechanisms by which kainic acid (KA) produces increases in [Ca++]i in single striatal neurons in vitro using fura-2-based microfluorimetry. When neurons were depolarized by perfusion with high K+ or veratridine containing solutions, [Ca++]i rose rapidly to a peak and then declined to a lower sustained plateau that persisted as long as the depolarizing stimulus. The peak high K+-induced rise in [Ca++]i occurred at [K+]o greater than 50 mM and the plateau was largest at 30 mM K+. [K+]o that was greater than 70 mM caused the magnitude of the plateau to decrease. Responses to high K+ stimulation were completely dependent on [Ca++]o and presumably represented Ca++ influx. Nitrendipine partially blocked the peak of the high K+-induced response and completely blocked the sustained plateau Ca++ influx. The nitrendipine-resistant portion of the high K+ response could be completely blocked by predepolarization of the cell in Ca++-free solution. KA also produced large increases in [Ca++]i that were abolished on removal of external Ca++. Predepolarization/nitrendipine greatly reduced the effect of lower [KA] (100 microM). However, KA-induced increases in [Ca++]i became increasingly resistant to block of voltage-sensitive Ca++ channels as [KA] rose above 100 microM, indicating a second route of Ca++ entry that may be the KA receptor-gated ionophore. About one-half the responses to KA (100 microM) also displayed a large oscillation. [Ca++]i rose to a peak, fell and then rose again before finally declining to a plateau level. This oscillation was abolished when all external Na+ was replaced by Li+ and may result from alterations in the buffering of [Ca++]i as a result of KA-induced Na+ influx.