AbstractIntrathoracic pressure (ITP) swings that permit spontaneous ventilation have physiological implications for the heart. We sought to determine the effect of respiration on cardiac output () during semi‐supine cycle exercise using a proportional assist ventilator to minimize ITP changes and lower the work of breathing (W_b). Twenty‐four participants (12 females) completed three exercise trials at 30%, 60% and 80% peak power (W_max) with unloaded (using a proportional assist ventilator, PAV) and spontaneous breathing. Intrathoracic and intraabdominal pressures were measured with balloon catheters placed in the oesophagus and stomach. Left ventricular (LV) volumes and were determined via echocardiography. Heart rate (HR) was measured with electrocardiogram and a customized metabolic cart measured oxygen uptake (). Oesophageal pressure swings decreased from spontaneous to PAV breathing by −2.8 ± 3.1, −4.9 ± 5.7 and −8.1 ± 7.7 cmH₂O at 30%, 60% and 80% W_max, respectively (P = 0.01). However, the decreases in W_b were similar across exercise intensities (27 ± 42 vs. 35 ± 24 vs. 41 ± 22%, respectively, P = 0.156). During PAV breathing compared to spontaneous breathing, decreased by −1.0 ± 1.3 vs. −1.4 ± 1.4 vs. −1.5 ± 1.9 l min⁻¹ (all P < 0.05) and stroke volume decreased during PAV breathing by −11 ± 12 vs. −9 ± 10 vs. −7 ± 11 ml from spontaneous breathing at 30%, 60% and 80% W_max, respectively (all P < 0.05). HR was lower during PAV breathing by −5 ± 4 beats min⁻¹ at 80% W_max (P < 0.0001). Oxygen uptake decreased by 100 ml min⁻¹ during PAV breathing compared to spontaneous breathing at 80% W_max (P < 0.0001). Overall, attenuating ITPs mitigated LV preload and ejection, thereby suggesting that the ITPs associated with spontaneous respiration impact cardiac function during exercise. imageKey points Pulmonary ventilation is accomplished by alterations in intrathoracic pressure (ITP), which have physiological implications on the heart and dynamically influence the loading parameters of the heart. Proportional assist ventilation was used to attenuate ITP changes and decrease the work of breathing during exercise to examine its effects on left ventricular (LV) function. Proportional assist ventilation with progressive exercise intensities (30%, 60% and 80% W_max) led to reductions in cardiac output at all intensities, primarily through reductions in stroke volume. Decreases in LV end‐diastolic volume (30% and 60% W_max) and increases in LV end‐systolic volume (80% W_max) were responsible for the reduction in stroke volume. The relationship between cardiac output and oxygen uptake is disrupted during respiratory muscle unloading.