To investigate and compare the potential of right to left ventricular volume (RVV/LVV) and diameter ratios (RVD/LVD) for the prediction of pulmonary arterial hypertension (PH) over 40 mmHg, and then to evaluate the incremental value of the pulmonary artery diameter index (PADi; PAD/body surface area) in the prediction of PH over 40 mHg. We correlated the followings on chest CT with same-day echocardiography-derived pulmonary arterial systolic pressure (PASP) in 139 patients (64.8 ± 15.4 years; 63 male): PADi, RVD/LVD on axial (RVD_axial/LVD_axial) and four chamber (RVD_4CH/LVD_4CH) views, and RVV/LVV. Those were then adjusted (multiplied) by PADi. Areas under the curves (AUC) for predicting PASP > 40 mmHg were calculated. All patients had undergone non-ECG-gated, contrast enhanced chest CT on a 64-slice multi-detector CT system. The correlation of PASP with RVV/LVV (R² = 0.45) was stronger than with RVD_axial/LVD_axial (R² = 0.28) and RVD_4CH/LVD_4CH (R² = 0.34). When adjusted by PADi, the correlation improved; R² = 0.55, 0.41 and 0.47, respectively. The AUC of adjusted RVV/LVV was significantly higher than those of other CT measurements (P < 0.05) excluding adjusted RVD_4CH/LVD_4CH (P = 0.08). With 18.80 as the cutoff, its sensitivity and specificity for predicting PH > 40 mmHg were 83 and 87 %, respectively. Ventricular volume is superior to ventricular diameter ratios in the prediction of PH > 40 mmHg with better correlation with PASP. PADi adds incremental value to these measurements in the prediction of PH > 40 mmHg. Of those, adjusted RVV/LVV is most reliable and predicts PH > 40 mmHg with fairly high sensitivity and specificity.