The purpose of this investigation was to validate in vivo the accuracy of videoangiocardiographic right ventricular volume determination. Thus the stroke volume of the right ventricle as difference of enddiastolic and endsystolic volume was determined simultaneously (n = 70) by videoangiocardiography and with an electromagnetic flowmeter in 7 open-chested pigs with an average weight of 18.5 kg. Model volumes, calculated with the multiple slices and the area-length methods on the basis of mono- and biplane angiographic right ventricular projections, were corrected conventionally and with factors appropriate for cardiac phase and spatial orientation. The correlation between electromagnetic and angiographic stroke volume was the lowest (r = 0.802) when monoplane and highest (r = 0.949) when biplane methods were used. A systematic error could be prevented (y = -0.3 + 1.0 x; Syx = +/- 1.96 ml) by applying correction factors appropriate for cardiac phase and spatial orientation. The statistical error was independent of the correction used. The correlation coefficient for the optimal corrected monoplane method based on the lateral projection was r = 0.861 and for the ap plane r = 0.930. Right ventricular stroke volume can be determined videoangiocardiographically without systemic error and with an acceptable statistical error, if spatial orientation and cardiac phase are considered. If only monoplane equipment is available, the approjection is preferable.