In order to characterize the day to day reproducibility of arrhythmias provoked during electrophysiologic stimulation, 114 patients with documented sustained clinical ventricular tachyarrhythmias were studied. Two baseline electrophysiologic tests were performed in the drug-free state and within 6 to 24 hours of one another. There was a significant increment (p less than or equal to 0.02) in the induction of sustained ventricular tachyarrhythmias as the number of programmed extrastimuli increased from one (10% induction) to four (64% induction). Provoked arrhythmias were observed to be more frequently nonreproducible (as reflected in a major change in rate or duration, or both, of an induced ventricular arrhythmia between baseline tests) as the number of extrastimuli increased from one (7%) to four (27%). Nonreproducibility with three and four extrastimuli was not significantly greater than when two extrastimuli were utilized. Electrophysiology-directed drug trials should be interpreted in light of this observed variability in induced arrhythmias.