Measurement of transepithelial potential difference (PD) on the nasal mucosa has been proposed to test for defective ion transport in cystic fibrosis (CF), and its possible correction after gene therapy or other treatments. The "classical" method records nasal PD under the inferior turbinate, with the disadvantage that the tip of the electrode is not seen by the operator. We have developed a purpose-designed perfusion electrode for PD recording on the visible, medial/posterior aspect of the turbinate. We wanted to determine whether such PD recordings adequately discriminate between CF patients and normal subjects. Measurements of baseline PD and response to a standardized perfusion protocol were performed in 20 normal subjects and 12 CF patients. Solutions of amiloride, with or without low chloride buffer were applied for 3 min. Increased baseline PD and depolarization after amiloride discriminated CF patients from normal subjects. Only one CF patient overlapped with the normal range. Superfusion of low chloride buffer with amiloride and terbutaline caused repolarization in 18 out of 20 normal subjects (90%), consistent with physiological Cl- secretion process, but in none of the CF patients. We conclude that measurements of potential difference on the medial/posterior aspect of the turbinate can discriminate between cystic fibrosis patients and normal subjects. At this site, visual control of the measurement is possible, and the mucosa is easily accessible for subsequent cytological sampling or biopsy.