Spontaneous sarcoplasmic reticulum (SR) Ca(2+) release causes delayed afterdepolarizations (DADs) via Ca(2+)-induced transient inward currents (I:(ti)). However, no quantitative data exists regarding (1) Ca(2+) dependence of DADs, (2) Ca(2+) required to depolarize the cell to threshold and trigger an action potential (AP), or (3) relative contributions of Ca(2+)-activated currents to DADs. To address these points, we evoked SR Ca(2+) release by rapid application of caffeine in indo 1-AM-loaded rabbit ventricular myocytes and measured caffeine-induced DADs (cDADs) with whole-cell current clamp. The SR Ca(2+) load of the myocyte was varied by different AP frequencies. The cDAD amplitude doubled for every 88+/-8 nmol/L of Delta[Ca(2+)](i) (simple exponential), and the Delta[Ca(2+)](i) threshold of 424+/-58 nmol/L was sufficient to trigger an AP. Blocking Na(+)-Ca(2+) exchange current (I(Na/Ca)) by removal of [Na](o) and [Ca(2+)](o) (or with 5 mmol/L Ni(2+)) reduced cDADs by >90%, for the same Delta[Ca(2+)](i). In contrast, blockade of Ca(2+)-activated Cl(-) current (I(Cl(Ca))) with 50 micromol/L niflumate did not significantly alter cDADs. We conclude that DADs are almost entirely due to I(Na/Ca), not I(Cl(Ca)) or Ca(2+)-activated nonselective cation current. To trigger an AP requires 30 to 40 micromol/L cytosolic Ca(2+) or a [Ca(2+)](i) transient of 424 nmol/L. Current injection, simulating I(ti)s with different time courses, revealed that faster I:(ti)s require less charge for AP triggering. Given that spontaneous SR Ca(2+) release occurs in waves, which are slower than cDADs or fast I(ti)s, the true Delta[Ca(2+)](i) threshold for AP activation may be approximately 3-fold higher in normal myocytes. This provides a safety margin against arrhythmia in normal ventricular myocytes.