A nonvolitional test to assess diaphragm strength in neonates has not been previously described. Our aim was to assess the feasibility of cervical (CMS) and anterior (AMS) magnetic stimulation of the phrenic nerves in neonates. Double circular stimulating coils (90-mm) were used. For CMS, one coil was placed over the cervical spine to bilaterally stimulate the phrenic nerve roots, whereas for AMS the coils were placed on the anterolateral aspect of the neck to allow unilateral and bilateral stimulation. Diaphragm contractility was assessed as transdiaphragmatic pressure (Pdi) measured with balloon catheters positioned in the midesophagus and stomach. Stimulus supramaximality was assessed by examining diaphragm twitch Pdi (TwPdi) across a range of stimulator outputs; 85, 90, 95, and 100% of maximum. Pressure signals were measured by differential pressure transducer and displayed in real time on a computer. Patients were studied supine during sleep. CMS was performed on seven neonates (mean gestational age [GA] 38 wk, range 33 to 40 wk) and AMS on 18 neonates (mean GA 37 wk, range 32 to 41 wk). The mean (SD) TwPdi with CMS was 2.5 (0.8) cm H(2)O. CMS was not supramaximal; reducing the stimulator output below 100% caused marked reductions in TwPdi, also the shape of the pressure waveforms suggested that CMS may not have activated the diaphragm alone. Mean (SD) TwPdi with AMS was 4.5 (1.3) cm H(2)O on the left, 4.1 (0.9) cm H(2)O on the right, and 8.7 (3.9) cm H(2)O for bilateral stimulation. The shape of the pressure waveforms suggested that AMS was more specific and a plateau in TwPdi at higher stimulator outputs indicated supramaximality. We conclude that AMS may provide a useful technique to assess diaphragm function in the neonate.