BACKGROUND: Inspiratory muscle weakness is a recognised cause of unexplained dyspnoea. It may be suggested by the finding of a low static inspiratory mouth pressure (MIP), but MIP is a difficult test to perform, with a wide normal range; a low MIP may also occur if the patient has not properly performed the manoeuvre. Further investigation conventionally requires balloon catheters to obtain oesophageal (Poes) and transdiaphragmatic pressure (Pdi) during sniffs or phrenic nerve stimulation. Two non-invasive tests of inspiratory muscle strength have recently been described--nasal pressure during a maximal sniff (Sn Pnas) and mouth pressure during magnetic stimulation of the phrenic nerves (Tw Pmo). The use of these two tests in combination might identify patients without inspiratory muscle weakness who are unable to produce a satisfactory MIP< therefore avoiding the need for investigation with balloon catheters. METHODS: Thirty consecutive patients with clinically suspected inspiratory muscle weakness and a low MIP underwent both conventional (Sn Poes and Tw Pdi) and non-invasive testing (Sn Pnas and Tw Pmo). Weakness was considered to be excluded by a Sn Poes of > or = 80 cm H20 or a Tw Pdi of > or = 20 cm H20. The limit values used to test the hypothesis were Sn Pnas > or = 70 cm H20 or Tw Pmo > or = 12 cm H20. RESULTS: Inspiratory muscle weakness was excluded in 17 of the 30 patients. Fifteen of these would have been identified using Sn Pnas and Tw Pmo, with better results when the two tests were combined. The cut off values selected for Sn Pnas and Tw Pmo were shown by ROC plots to indicate normal strength conservatively, avoiding failure to detect mild degrees of weakness. No patient with global weakness was considered normal by Sn Pnas or Tw Pmo. CONCLUSIONS: In most patients with normal inspiratory strength and a low MIP, Tw Pmo and Sn Pnas used in combination can reliably exclude global inspiratory muscle weakness, reducing the number of patients who need testing with balloon catheters.