Brief supramaximal stimulation of the phrenic nerves (twitch) is considered a promising technique to detect diaphragmatic fatigue in humans. However, the most commonly employed methodology (transcutaneous stimulation) is technically difficult. Cervical magnetic stimulation is a recently described technique that is potentially simpler and may obviate some of the problems inherent with transcutaneous stimulation. The purpose of this study was to determine the ability of cervical magnetic stimulation to evaluate diaphragmatic function. Accordingly, we measured transdiaphragmatic pressure (Pdi) during transcutaneous and cervical magnetic stimulation of the phrenic nerves before and after a potentially fatiguing task; inspiratory threshold loading to task failure. During threshold loading, subjects generated approximately 60% of their maximal esophageal pressure with each breath until they could no longer reach the target pressure. At least 10 twitches were obtained during both transcutaneous and magnetic stimulation before and 10, 30, 60, and 120 min after threshold loading. Control twitch Pdi was significantly larger during magnetic stimulation compared with transcutaneous stimulation: 39.3 +/- 3.0 (mean +/- SE) versus 27.4 +/- 2.3 cm H2O, p < 0.0005. This increase in twitch Pdi was solely due to the esophageal component. Following threshold loading, a significant reduction in transcutaneous twitch Pdi was seen in only three of the 10 subjects. Mean transcutaneous twitch Pdi fell only slightly from 27.4 +/- 2.3 during control to 25.1 +/- 2.2 cm H2O at 10 min after loading (p < 0.004). In contrast, magnetic twitch Pdi was significantly reduced in nine of the 10 subjects following threshold loading. Mean magnetic twitch Pdi fell from 39.3 +/- 3.0 during control to 31.1 +/- 3.0 cm H2O at 10 min after loading (p < 0.0001). The average fall in twitch Pdi post-loading (expressed as a percentage of the control value) was significantly greater for magnetic stimulation compared with transcutaneous stimulation: 21.0 +/- 3.1 versus 7.8 +/- 2.9%, p < 0.0001. In summary: (1) in the fresh state, twitch Pdi is larger with magnetic stimulation compared with transcutaneous stimulation, and (2) transcutaneous and cervical magnetic twitch Pdi are affected differently by threshold loading to task failure.