Determining ventricular capture when using transcutaneous pacing (TCP) can be difficult. Our objective was to evaluate the use of ultrasound (US) to determine capture in patients undergoing TCP. We studied 15 patients in a convenience sample. Videotaped US views were obtained during TCP with concurrent electrocardiographic monitoring on the US screen. The treating physician made capture determinations first by using clinical indicators, followed by US images. Two reviewing physicians blinded to the treating physician's findings and to each other determined capture on videotape. Kappa levels of agreement were calculated. A value of kappa = 0.84 was found for agreement of the treating physician's findings through clinical measures and US. Agreement between the first reviewer and the treating physician had kappa =.47 and between the second reviewer and the treating physician was had kappa =.67. All P values were <.05. US and clinical impression of ventricular capture have good to excellent agreement. US may be useful for determining ventricular capture in patients undergoing TCP.