BACKGROUND Preliminary evidence suggests that regular inhalation of nebulised amiloride reduces sputum viscoelasticity, increases the clearance of sputum by mucociliary mechanisms and by coughing and reduces the rate of deterioration in lung function in patients with cystic fibrosis. These effects depend on adequate delivery of amiloride to the airways. This study was performed to quantify and compare pulmonary deposition of amiloride produced by two different nebuliser systems. METHODS The pulmonary deposition of nebulised amiloride (1 mg in 3 ml saline) was measured in eight patients with cystic fibrosis when given via a jet (System 22 with CR 60 compressor) and an ultrasonic (Fisoneb) nebuliser. Human serum albumin labelled with technectium-99m was used as an indirect marker for amiloride and its deposition in the lung was detected with a gamma camera. RESULTS Amiloride inhalation caused no side effects or changes in spirometric indices. The mean (SD) total pulmonary amiloride deposition was 57 (24) micrograms with the System 22 and 103 (53) micrograms with the Fisoneb nebuliser. Pulmonary deposition was completed more rapidly with the Fisoneb (4-5 minutes) than with the System 22 nebuliser (7-8 minutes) and the Fisoneb was preferred by the patients. CONCLUSIONS Both nebulisers appeared to deliver adequate amounts of amiloride to the lungs, but treatment with the Fisoneb nebuliser was quicker, more efficient, and more acceptable to the patients. Of the two nebulisers assessed, the Fisoneb would be preferred for clinical trials.