Assessing left ventricular (LV) filling pressure (pulmonary capillary wedge pressure, PCWP) is an important aspect in the care of patients with heart failure (HF). Physicians rely on right ventricular (RV) filling pressures such as central venous pressure (CVP) to predict PCWP, assuming concordance between CVP and PCWP. However, the use of this method is limited because discordance between CVP and PCWP is observed. We hypothesized that PCWP can be reliably predicted by CVP corrected by the relationship between RV and LV function, provided by the ratio of tissue Doppler peak systolic velocity of tricuspid annulus (S _T) to that of mitral annulus (S _M) (corrected CVP:CVP·S _T/S _M). In 16 anesthetized closed-chest dogs, S _T and S _M were measured by transthoracic tissue Doppler echocardiography. PCWP was varied over a wide range (1.8–40.0 mmHg) under normal condition and various types of acute and chronic HF. A significantly stronger linear correlation was observed between CVP·S _T/S _M and PCWP (R ² = 0.78) than between CVP and PCWP (R ² = 0.22) (P < 0.01). Receiver-operating characteristic (ROC) analysis indicated that CVP·S _T/S _M >10.5 mmHg predicted PCWP >18 mmHg with 85 % sensitivity and 88 % specificity. Area under ROC curve for CVP·S _T/S _M to predict PCWP >18 mmHg was 0.93, which was significantly larger than that for CVP (0.66) (P < 0.01). Peripheral venous pressure (PVP) corrected by S _T/S _M (PVP·S _T/S _M) also predicted PCWP reasonably well, suggesting that PVP·S _T/S _M may be a minimally invasive alternative to CVP·S _T/S _M. In conclusion, our technique is potentially useful for the reliable prediction of PCWP in HF patients.