Hyperventilation and inadequate cardiac output (CO) increase are the main causes of exercise limitation in pulmonary hypertension (PH). Intrapulmonary blood flow partitioning between ventilated and unventilated lung zones is unknown. Thoracic impedance cardiography and inert gas rebreathing have been both validated in PH patients for non-invasive measurement of CO and pulmonary blood flow (PBF), respectively. This study sought to evaluate CO behaviour in PH patients during exercise and its partitioning between ventilated and unventilated lung areas, in parallel with ventilation partitioning between ventilated and unventilated lung zones. Eighteen PH patients (group 1 or 4) underwent a cardiopulmonary exercise test (CPET) with a three-step loaded workload protocol. The steps occurred at 0%, 20%, 40%, and 60% of peak workload reached during a preliminary maximum CPET. Ventilatory parameters, arterial blood gases, CO, PBF, and intrapulmonary shunt (calculated as the difference between CO and PBF) were obtained at each step, combining thoracic impedance cardiography and an inert gas rebreathing technique. Dead space ventilation observed throughout the exercise was about 40% of total ventilation. A progressive increase of CO from 4.86 ± 1.24 L/min (rest) to 9.41 ± 2.63 L/min (last step), PBF from 3.81 ± 1.41 L/min to 7.21 ± 2.93 L/min, and intrapulmonary shunt from 1.05 ± 0.96 L/min to 2.21 ± 2.28 L/min was observed. Intrapulmonary shunt was approximately 20% of CO at each exercise step. Although the study population was small, the combined non-invasive CO measurement seems a promising tool for deepening our knowledge of lung exercise haemodynamics in PH patients. This technique could be applied in future studies to evaluate PH treatment influences on CO partitioning, since a secondary increase of intrapulmonary shunt is undesirable.