BACKGROUND: Static vascular filling pressures suffer from poor predictive power in identifying the volume-responsive heart. The use of dynamic arterial pressure variables, including pulse pressure variation (PPV) has instead been suggested to guide volume therapy. The aim of the present study was to evaluate the performance of several clinically applicable haemodynamic parameters to predict volume responsiveness in a pig closed chest model of acute left ventricular myocardial infarction. METHODS: Fifteen anaesthetized, mechanically ventilated pigs were studied following acute left myocardial infarction by temporary coronary occlusion. Animals were instrumented to monitor central venous (CVP) and pulmonary artery occlusion (PAOP) pressures and arterial systolic variations (SPV) and PPV. Cardiac output (CO) was measured using the pulmonary artery catheter and by using the PiCCO monitor also giving stroke volume variation (SVV). Variations in the velocity time integral by pulsed-wave Doppler echocardiography were determined in the left (DeltaVTI(LV)) and right (DeltaVTI(RV)) ventricular outflow tracts. Consecutive boluses of 4 ml/kg hydroxyethyl starch were administered and volume responsiveness was defined as a 10% increase in CO. RESULTS: Receiver-operator characteristics (ROC) demonstrated the largest area under the curve for DeltaVTI(RV) [0.81 (0.70-0.93)] followed by PPV [0.76 (0.64-0.88)] [mean (and 95% CI)]. SPV, DeltaVTI(LV) and SVV did not change significantly during volume loading. CVP and PAOP increased but did not demonstrate significant ROC. CONCLUSION: PPV may be used to predict the response to volume administration in the setting of acute left ventricular myocardial infarction.