INTRODUCTION: During the coronavirus disease 2019 pandemic, public health officials widely adopted the use of face masks (FM) to minimize infections. Despite consistent evidence that FMs increase dyspnea, no studies have examined the multidimensional components of dyspnea or their underlying physiological mechanisms. METHODS: In a randomized crossover design, 16 healthy individuals ( n = 9 women, 25 ± 3 yr) completed incremental cycling tests over three visits, where visits 2 and 3 were randomized to either surgical FM or no mask control. Dyspnea intensity and unpleasantness were assessed throughout exercise (0-10 Borg scale), and the Multidimensional Dyspnea Profile was administered immediately after exercise. Crural diaphragmatic EMG and esophageal pressure were measured using a catheter to estimate neural respiratory drive and respiratory muscle effort, respectively. RESULTS: Dyspnea unpleasantness was significantly greater with the FM at the highest equivalent submaximal work rate achieved by a given participant in both conditions (iso-work; 5.9 ± 1.7 vs 3.9 ± 2.9 Borg 0-10 units, P = 0.007) and at peak exercise (7.8 ± 2.1 vs 5.9 ± 3.4 Borg 0-10 units, P = 0.01) with no differences in dyspnea intensity ratings throughout exercise compared with control. There were significant increases in the sensory quality of "smothering/air hunger" ( P = 0.01) and the emotional response of "anxiousness" ( P = 0.04) in the FM condition. There were significant increases in diaphragmatic EMG and esophageal pressure at select submaximal work rates, but no differences in heart rate, pulse oximetry-derived arterial oxygen saturation, or breathing frequency throughout exercise with FMs compared with control. FMs significantly reduced peak work rate and exercise duration (both P = 0.02). CONCLUSIONS: FMs negatively impact the affective domain of dyspnea and increase neural respiratory drive and respiratory muscle effort during exercise, although the impact on other cardiorespiratory responses are minimal.