A method and apparatus for validating cardiac pacing pulse train preceding a T- shock. Validation of the pacing pulse train includes verifying capture of at least the last pacing pulse of the pulse train and verifying that intrinsic ventricular events are not sensed during a pacing train interval that includes at least the interval between the last pacing pulse and a scheduled T-shock delivery. Capture verification may include sensing an event during an ER (evoced response) sensing window, morphological analysis of a sensed event, and analysis of the temporal relationship between sensed events occurring on different EGM sources. The scheduled T-shock is delivered in response to a valid pacing pulse train. A response to an invalid pacing pulse train may include any of an invalid pacing train notification, withholding a scheduled T-shock, extension of the pacing train, repeating the pacing train, or adjusting the pacing pulse train parameters.