Chronic pulmonary infection is the major cause of morbidity and mortality in cystic fibrosis. High levels of DNA in the sputum make the sputum viscous and difficult to expectorate. Recombinant human deoxyribonuclease (rhDNase) in vitro has been shown to reduce the viscoelasticity of the sputum from CF patients. We have done a phase II double-blind randomised placebo-controlled trial in which patients received either 2.5 mg rhDNase twice daily or placebo for 10 days. All patients had forced vital capacity (FVC) above 40% predicted and were clinically stable. Patients were followed up for 42 days from the start of drug/placebo administration. All 71 randomised patients, aged 16-55, completed every aspect of the study and baseline characteristics were similar in the two groups. Baseline forced expiratory volume in one second (FEV1) was 46% of predicted for patients randomised to rhDNase, and 48% for those randomised to placebo; and baseline FVC was 76% of predicted for both groups. The mean percentage change in FEV1 from baseline was a 13.3% rise on rhDNase and a 0.2% fall on placebo (p < 0.001). FVC rose 7.2% in the rhDNase group and 2.3% in the placebo group (not significant). There were no life-threatening adverse events and no anaphylactic reactions. There was no significant difference in side-effects between the groups. This study confirms that short-term administration of rhDNase in stable patients with cystic fibrosis is safe and improves lung function.