1. Using the patch-clamp technique we have investigated the kinetic and pharmacological properties of high-voltage-activated (HVA) Ca2+ channels in short-term-cultured cat chromaffin cells. 2. In 10 mM Ba2+, HVA currents activated around -40 mV, reached maximal amplitude at 0 mV and reversed at about +60 mV. At 0 mV, HVA current activation was fast (mean tau act, 2.45 ms), and followed by either an incomplete inactivation or by a second slow phase of activation (mean tau slow, 36.8 ms) that was lost when Ba2+ was replaced by Ca2+. HVA Ba2+ currents deactivate quickly on repolarization to -50 mV (mean tau deact, 0.36 ms). 3. In most cells, HVA currents were sensitive to common dihydropyridine (DHP) derivatives. Nisoldipine blocked the currents maximally at low membrane potentials (mean block 76% at -30 mV, 3 microM) and gradually less at higher voltages. Nisoldipine block was clearly time dependent (33 and 56% after 30 and 600 ms, respectively, to 0 mV). 4. Bay K 8644 (3 microM) action was variable and caused (1) a 2- to 4-fold increase of Ba2+ currents at -40 to -20 mV, (2) a -15 mV shift of the current-voltage relationship and (3) a 10- to 20-fold prolongation of HVA channel deactivation at -50 mV. 5. Nisoldipine block and Bay K 8644 potentiation of HVA currents increased markedly in omega-conotoxin GVIA (omega-CgTX)-pretreated cells, suggesting an increased fraction of DHP-sensitive currents in these cells. Nisoldipine block of residual omega-CgTX-resistant currents was almost complete (mean block, 82%) during pulses of 1 s to 0 mV. 6. The degree of inhibition produced by omega-CgTX (2 microM for 1 min) varied from cell to cell (mean block, 46%) and was partly reversible. Residual omega-CgTX-resistant currents exhibited faster activation-deactivation kinetics than control currents. 7. The slow phase of HVA current activation was abolished if a conditioning prepulse of 40 ms to +70 mV preceded a test pulse to 0 mV. Double-pulse protocols caused an average current increase (facilitation) of 37% that was voltage dependent and which correlated with the slow phase of Ca2+ channel activation. Facilitation was lost in most omega-CgTX-treated cells and was little affected by nisoldipine (3 microM) and Bay K 8644 (1 microM). Facilitation was potentiated in cells dialysed with 100 microM guanosine 5'-O-(3-thiotriphosphate) (GTP-gamma-S) and fully prevented by 1 mM guanosine 5'-O-(2-thiodiphosphate) (GDP-beta-S).(ABSTRACT TRUNCATED AT 400 WORDS)