The finding of hemidiaphragm elevation on a chest radiograph, in absence of an ipsilateral lung disease, is assumed to indicate severe hemidiaphragm dysfunction. To test this hypothesis we retrospectively reviewed chest radiograph findings and corresponding twitch transdiaphragmatic pressure (T(W)P(DI)) results from 42 (17 female, age range 22-79 years) consecutive patients who underwent phrenic nerve stimulation studies. Chest radiographs were independently reviewed in a blind manner by two radiologists. The interobserver agreement was moderate, the kappa value ranging from 0.48 (left hemidiaphragm) to 0.59 (lung parenchyma). Hemidiaphragm dysfunction was diagnosed if T(W)P(DI) of corresponding hemidiaphragm was less than 3.5 cm H2O. The prevalence of patients with an elevated unilateral hemidiaphragm on chest radiograph was 64% and of patients with unilateral paralysis judged by T(W)P(DI) was 24%. Sensitivity, specificity, positive and negative predictive values for chest radiograph, as a diagnostic test for unilateral diaphragm dysfunction were 0.90, 0.44, 0.33 and 0.93, respectively. We conclude that the isolated elevation of hemidiaphragm on chest radiograph is of little value in the diagnosis of unilateral hemidiaphragm paralysis, though the condition is unlikely if diaphragm elevation is absent.