1. Low voltage-activated (LVA) Ca2+ current in clonal (GH3) pituitary cells was studied with the use of the whole-cell recording technique. The use of internal fluoride to facilitate the rundown of high voltage-activated (HVA) Ca2+ current allowed the study of LVA current in virtual isolation. 2. In 10 mM [Ca2+]o, detectable LVA current begins to appear at about -50 mV, with half-maximal activation occurring at -33 mV. The time course of activation was best described by a Hodgkin-Huxley expression with n = 3, suggesting that at least three closed states must be traversed before channel opening. 3. Deactivation was found to vary exponentially with membrane potential between -60 and -160 mV, indicating that channel closing is rate-limited by a single, voltage-dependent transition. 4. Onset and removal of inactivation between -40 and -130 mV were best described by the sum of two exponentials. Between -80 and -130 mV, both components of removal of inactivation showed little voltage dependence, with time constants of approximately 200-300 ms and 1-2 s. At membrane potentials above -40 mV, a single component of inactivation onset was detected. This component was voltage independent between -20 and +20 mV (tau = 22 ms). Thus inactivation of LVA current is best described by multiple, voltage-in-dependent processes. 5. Significant inactivation of LVA current occurred at -65 mV without detectable macroscopic current. This suggests that inactivation is not strictly coupled to channel opening. 6. Peak LVA current increased with increasing [Ca2+]o, with saturation approximately 50 mM. The Ca(2+)-dependence of peak LVA current was reasonably well described by a single-site binding isotherm with half-maximal LVA current at approximately 7 mM. 7. LVA current in GH3 cells was largely resistant to blockade by Ni2+. The relative potency of inorganic cations in blocking GH3 LVA current was (concentrations which produced 50% block): La3+ (2.4 microM) greater than Cd2+ (188 microM) greater than Ni2+ (777 microM). 8. Several organic agents, including putative LVA blockers, HVA current blockers and various anesthetic agents, were tested for their ability to block LVA current. The concentrations that produced 50% block are as follows: nifedipine (approximately 50 microM), D600 (51 microM), diltiazem (131 microM), octanol (244 microM), pentobarbital (985 microM), methoxyflurane (1.41 mM), and amiloride (1.55 mM). Phenytoin and ethosuximide produced 36 and 10% block at 100 microM and 2.5 mM, respectively.(ABSTRACT TRUNCATED AT 400 WORDS)