Recent studies showed that regional pulmonary perfusion can be reliably estimated using electrical impedance tomography (EIT) with the aid of hypertonic saline based contrast enhancement. Building on these successful studies, we studied contrast EIT for pulmonary perfusion defect caused by an artificially induced pulmonary embolism (PE) in a large ovine model (N = 8, 78 ± 7.8 kg). Furthermore, the efficacy of a less invasive contrast bolus of 0.77 ml kg(-1) of NaCl 3% was compared with a more concentrated bolus of 0.13 ml kg(-1) of NaCl 20%. Prior to the injection of each contrast bolus injection, ventilation was turned off to provide a total of 40 to 45 s of apnoea. Each bolus of impedance contrast was injected through a catheter into the right atrium. Pulmonary embolisation was performed by balloon occlusion of part of the right branch of the pulmonary trunk. Four parameters representing the kinetics of the contrast dilution in the lung were evaluated for statistical differences between baseline and PE, including peak value, maximum uptake, maximum washout and area under the curve of the averaged contrast dilution curve in each lung. Furthermore, the right lung to left lung (R2L) ratio of each the aforementioned parameters were assessed. While all of the R2L ratios yielded significantly different means between baseline and PE, it can be concluded that the R2L ratios of area under the curve and peak value of the averaged contrast dilution curve are the most promising and reliable in assessing PE. It was also found that the efficacy of the two types of impedance contrasts were not significantly different in distinguishing PE from baseline in our model.