It is generally accepted that the plain chest roentgenogram is useful in evaluating the status of pulmonarycirculation. But it is conflicting concerning the accuracy of this method for estimating pulmonary blood flow,resistance and pressure. Most studies have been performed in mitral stenosis. However, out present study dealswith 97 cases of congenital heart disease with left to right shunt including 22 cases of ASD, 48 cases of VSD and32 cases of PDA. To determine the accuracy of roentgenologic evaluation of hemodynamic status and localization ofshunt level, the findings of plain chest film are compared with the datas of cardiac catheterization in each caseof congenital heart disease. The results are summarised as follows ; 1. When the pulmonary vascular resistance isnormal, there is relative correlation between the degree of pulmonary plethora and actual shunt volume, but notwell-correlated quantitatively. The degree of correlation is different depending on the disease entity; with thesame volume of shunt the pulmonary plethora is most prominent in VSD, and is least in ASD. The pulmonary plethoraitself dose not give to assist in determining the location of shunt. 2. When pulmonary vascular resistance isincreased, the degree of pulmonary plethora is not correlated at all to the shunt volume. The presence ofcentralization of pulmonary vascularity in plain chest is a good index of increased pulmonary vascular resistance.Centralization is present in all cases whose pulmonary vascular pattern is detectable in 13 cases among 32 casesof the subjects having increased pulmonary vascular resistance in this study. The detection rate is not apparentlydifferent depending on disease entities. This poor result is a big barrier in this study; the cases havingincreased pulmonary vascular resistance constitutes the major portion of the error in estimating hemodynamicchange from the plain film, since the pulmonary plethora is not correlated to shunt volume. 3. Pulmonary arterialpressure is presumptively elevated when the centralization of the pulmonary vascular pattern is present, however,in the remainder the radiologic criteria of pressure determination cannot be found. 4. The cross-sectional area ofright descending pulmonary artery corrected by body surface are is best correlated with pulmonary blood flow inASD having within normal range of vascular resistence (Y=0.013X+0.332, r=0.76, p<0.02, X=cross-sectional area ofRDPA/BSAm, Y=Q'p/Q's). The subjects having increased pulmonary vascular resistance have the right descendingpulmonary artery always larger in diameter than those who have not increased pulmonary vascular resistance in thecorresponding shunt volume. In PDA the relation is also well correlated but not as well as in the relation in ASD(Y=0.22X+0.224, r+0.717, p<0.01). With increase of pulmonary vascular resistance, the size of the right descendingpulmonary artery has no constant relation to those of normal pulmonary vascular resistance. There is nostatistically significant correlation in this relationship in subjects of VSD (Y=0.003X+1.390, r=0.17, p>0.1).