The myocardial performance index (MPI) reflects both the systolic and diastolic function of the heart, and is easily applied in practice. In this study, we aimed to determine the relationship between MPI and invasive haemodynamic parameters in heart failure patients. A total of 126 patients with heart failure were selected, all of whom were referred for diagnostic cardiac catheterisation, and were divided into two groups. Group I consisted of 59 patients (32 men and 27 women, mean age 61 +/- A 10; functional capacity New York Heart Association (NYHA) Class I; and left ventricular end-diastolic pressure (LVEDP) < 16 mmHg). Group II included 67 patients (34 men and 33 women, mean age 60 +/- A 9; NYHA Class a parts per thousand yen II; LVEDP a parts per thousand yen 16 mmHg). The following parameters were measured in all patients: ejection fraction with Simpson method, the peak mitral early (E) and late (A) diastolic velocities, E/A ratio, deceleration time (DT) and tissue Doppler from four different areas of the mitral annulus (septum, lateral, inferior and anterior). In order to measure MPI with two methods (standard Doppler and tissue Doppler), isovolumetric contraction time (IVCT), isovolumetric relaxation time (IVRT) and ejection time (ET) were measured from four areas and mean values of MPI were calculated. There was no difference between the two groups in E/A ratios, DT and IVRT (p > 0.05). Group II patients had longer IVCT and ET, when compared with group I patients (p < 0.05). MPI, measured by both standard pulsed wave Doppler and tissue Doppler methods, was significantly higher in group II patients, when compared with the values obtained from group I patients (Group I: 0.50 +/- A 0.2 and 0.50 +/- A 0.14; group II: 0.98 +/- A 0.3 and 1.2 +/- A 0.32; p < 0.001). According to receiver operating characteristics curve analysis, the cut-off value for MPI measured by tissue Doppler was 0.74. The sensitivity and specificity of this value were measured as 92.5 and 91.5 %, respectively. MPI measured by standard Doppler method was 0.67, and its sensitivity and specificity were 85.1 and 83.1 %, respectively. We found a strong relationship between MPI and LVEDP (r = 0.83, p < 0.001; r = 0.96, p < 0.001), especially when measured by tissue Doppler. In addition, we observed a significant relationship between the MPI values measured by tissue Doppler and those measured by standard traditional methods (r = 0.85, p < 0.001). We showed that MPI was reliable for the evaluation of global cardiac functions in patients with heart failure, as measured with both pulsed-wave Doppler and tissue Doppler. We assert that, in order to differentiate between those patients with symptomatic heart failure from the asymptomatic cases, MPI as measured with the tissue Doppler method is an improvement on MPI as measured using traditional methods.