1. Intracellular recording in the in vitro slice preparation and whole-cell, patch-clamp recording of acutely dissociated neurons from the rat lateral geniculate nucleus (LGN) were combined to study the Ca currents underlying their electrical responses. In slices from young animals (postnatal days 13-16), we found that dorsal LGN neurons have responses similar to those of adult preparations, including the presence of a low-threshold Ca spike (LTS). After enzymatic isolation of LGN neurons from the same animals, the firing properties appeared well preserved, as indicated by whole-cell, current-clamp recordings from dissociated multipolar cells (presumably geniculocortical relay neurons). 2. Two types of Ca currents were identified in voltage-clamped, isolated LGN neurons on the basis of their voltage dependency, pharmacology, and selectivity properties. These two currents resemble the low-voltage-activated (LVA) and high-voltage-activated (HVA) Ca channels found in rat sensory neurons (9). 3. The LVA current component required negative potentials (less than -80 mV) to deinactivate completely, started to activate around -60 mV and reached a plateau level around -25 mV. It peaked within 30-6 ms and decayed with a single time constant of approximately 24 ms at -20 mV. Its inactivation curve ranged from -100 to -40 mV, with a half-inactivation near -60 mV. The HVA current component could be isolated by holding the membrane potential positive to -60 mV, activated at potentials positive to -30 mV and peaked around +5 mV. The time-to-peak ranged from 30 to 6 ms in the voltage range from -30 to +35 mV and decayed very slowly with sustained depolarizing pulses (time constant ranged between 1,600 and 40 ms over the same voltage range). 4. The inactivation of LVA Ca current during depolarizing voltage steps was consistent with a voltage-dependent process. The recovery from inactivation after short (100 ms), inactivating prepulses displayed two exponential phases. The slower phase was predominant under conditions that induce large current flow through the membrane, suggesting a Ca-mediated mechanism. 5. The LVA current was preferentially blocked by 50 microM Ni2+, leaving the HVA currents almost unaltered. Fifty micromolars Cd2+, in contrast, seemed more effective in blocking the HVA component of the Ca current.(ABSTRACT TRUNCATED AT 400 WORDS)