In our clinical work we have occasionally encountered difficulties (e.g., no response, concomitant brachial plexus stimulation) in performing phrenic nerve conduction studies. The aim of this study was to overcome these difficulties and obtain our own set of normative data. In 29 healthy volunteers (15 men), aged 21-65 years, phrenic nerve conduction studies were performed using bipolar surface stimulation electrodes and a standard recording montage. Stimulation just above the clavicle, between the sternal and clavicular heads of the sternocleidomastoid muscles, elicited responses at the lowest stimulation strength, without concomitant brachial plexus stimulation. M-wave amplitude and duration changed with respiration, whereas latency and area did not. The normative limit for M-wave latency was 8.0 ms (upper), for amplitude it was 0.46/0.33 mV (lower: inspiration/expiration), and for area it was 4.4 mVms (lower). We suggest a slight modification of the generally used position for phrenic nerve stimulation, and the use of M-wave latency and area (unaffected by the respiratory cycle) in future phrenic nerve conduction studies.