1. Perforated patch recording was used to examine the influence of the calcium-dependent chloride current (iCl(Ca)) on Ca2+ action potentials in AtT-20 pituitary cells. The calculated chloride equilibrium potential (ECl) was adjusted by changing either intracellular or extracellular [Cl-]. Action potential duration varied as a function of ECl. When ECl was set at -21 mV, both spontaneous and evoked action potentials displayed a long plateau phase between -20 and -25 mV, which typically lasted for several seconds. Setting ECl to more negative potentials resulted in briefer action potentials; at an ECl of -52 mV, no plateau phase was evident. Spontaneous depolarization and action potential firing still occurred when ECl was negative to firing threshold, which indicates that the slow depolarizing wave that precedes the firing of spontaneous action potentials does not require activation of ICl(Ca). 2. In voltage clamp experiments the magnitude of ICl(Ca) diminished slowly during a prolonged depolarization, over a time course that coincided with action potential termination. 3. Niflumic acid (100 microM) blocked ICl(Ca) by 90% but had no effect on either K+ or Ca2+ currents. This concentration of niflumic acid eliminated the plateau phase, but did not prevent the firing, of Ca2+ action potentials. 4. Internal [Ca2+] was measured photometrically after loading cells with the Ca2+ indicator dye, Fura-2. Under voltage clamp conditions, concentrations of niflumic acid (30-100 microM) that blocked depolarization-evoked ICl(Ca) had little or no effect on simultaneously recorded Ca2+ transients. Perforated patch recording from Fura-loaded cells showed that action potentials were temporally associated with transient increases in intracellular [Ca2+]. Niflumic acid (30-100 microM) disrupted the rhythmic firing of spontaneous action potentials and associated intracellular Ca2+ transients. 5. Fluorescent measurements of Ca2+ transients were also made in cells unperturbed by patch recording, and were used as a measure of action potential duration in the absence of experimental alteration of internal [Cl-]. Spontaneous Ca2+ transients were of long duration (approximately 2 s), which suggests that intracellular [Cl-] is relatively high (40-50 mM) in these cells. The spontaneous Ca2+ transients were inhibited by niflumic acid. 6. Niflumic acid up to 100 microM, had neglible effects on either basal or stimulated (by 2 microM-(+/-)-isoprenaline) hormone secretion, as shown by radioimmunoassay of adrenocortotrophic hormone release.(ABSTRACT TRUNCATED AT 400 WORDS)