Assessing diaphragm function is clinically and physiologically pertinent. It can rely on the measurement of pressure responses to phrenic stimulation. Combining mouth pressure (Pm) with cervical magnetic stimulation (CMS) is painless and easy to perform, but Pm-CMS poorly reflects esophageal pressure (Pes-CMS) because of poor pressure transmission across the airway. We reasoned that the mouth opening and neck flexion that are associated with the measurement of Pm-CMS would impair upper airway dynamics and further hinder pressure transmission. Therefore, we assessed the CMS-related pressure measured in a nasal mask (Pmask; mouth closed) without neck flexion as a possible surrogate of Pes-CMS, in 14 men and 3 women, age 24.5+/-2.2. Pes-CMS was 15.7+/-4.3 cmH2O, significantly higher than Pm-CMS (13.5+/-5.6 cmH2O, P<0.0001) but not different from Pmask-CMS (15.2+/-4.9 cmH2O). The concordance correlation coefficient was low (0.6808) between Pes-CMS and Pm-CMS. It was higher between Pes-CMS and Pmask-CMS (0.8730). Pm-CMS wrongly classified five subjects as abnormal (<10 cmH2O), versus 1 for Pmask and 5 for Pm (P=0.025). Passing and Bablok regressions found no difference between Pes-CMS and Pmask-CMS, but identified a systematic difference and a proportional error between Pes-CMS and Pm-CMS. We conclude that Pmask-CMS is a better surrogate of Pes-CMS than Pm-CMS.