BACKGROUND: Osmotic agents, such as inhaled dry powder mannitol, may increase mucociliary clearance by rehydrating the airway surface liquid and thus act as disease-modifying treatments in cystic fibrosis (CF). This is the first therapeutic trial of inhaled mannitol in children with CF; it was compared with recombinant human deoxyribonuclease (rhDNase), the current best established mucolytic treatment. METHODS: 38 children were recruited to an open crossover study. Subjects underwent an initial bronchial provocation challenge with dry powder mannitol. Those children with a negative challenge were randomly allocated to one of three consecutive 12-week treatment blocks (inhaled mannitol alone, nebulised rhDNase alone and mannitol + rhDNase). The primary outcome was forced expiratory volume in 1 s (FEV(1)). A number of secondary outcome measures were also studied. RESULTS: Twenty children completed the study. Bronchoconstriction and cough associated with mannitol administration contributed to the high attrition rate. The mean increase in FEV(1) following 12 weeks of treatment was 0.11 litres (6.7%) (p = 0.055) for mannitol alone, 0.12 litres (7.2%) (p = 0.03) for rhDNase alone and 0.03 litres (1.88%) (p = 0.67) for rhDNase and mannitol. None of the secondary clinical outcomes was statistically significantly different between treatments. CONCLUSIONS: Inhaled mannitol was at least as effective as rhDNase after 3 months treatment. There was a marked individual variation in tolerance to mannitol and in response to treatment however. Children who do not respond to rhDNase many benefit from a trial of inhaled mannitol. The combination of mannitol and rhDNase was not useful.