The objective of this study was to compare the long-term safety of a fixed combination of fenoterol hydrobromide (50 microg) and ipratropium bromide (20 microg) delivered using a metered dose inhaler (MDI) formulated with a non-chlorinated propellant, hydrofluoroalkanel34a (HFA-MDI), with delivery using the conventional chlorofluorocarbon propellant (CFC-MDI, Berodual/Bronchodual). The study was designed according to Safety Assessment of Marketed Medicines (SAMM) guidelines, to reflect as far as possible the use of MDls under normal prescribing conditions. Two thousand and twenty-seven patients with chronic airways obstruction (CAO) were enrolled from 99 centres in France, 95 centres in Germany and 24 centres in Italy. Following a 2-week run-in period, patients were randomized on a 2:1 basis (1,348 patients to HFA-MDI, 679 patients to CFC-MDI) to receive a flexible dose regimen of the combination (2 puffs, 2-4 times a day, as prescribed by the investigator) during a 12-week open label phase. The overall incidence of adverse events was comparable between both groups. In addition, the incidence of respiratory side effects was also similar, with CAO exacerbations or bronchitis the most frequently recorded events. The safety profile of the HFA formulation was comparable to those of the marketed CFC-MDIs used in Germany and France/Italy. No clinically significant differences were detected between HFA134a or CFC driven inhalers on the switch from CFC- to HFA-MDI (2 weeks before randomisation versus 2 weeks after randomization). There was a trend for taste complaints to be reported more frequently by patients in the HFA-MDI group (0.7% before randomization versus 3.4% after randomization). This, however, was an expected finding as the HFA134a formulation does have a different taste to the CFC formulation. No difference between formulations was observed in the incidences of coughing or paradoxical bronchospasm. The incidence of falls in FEV1 >15% within 15 min following inhalation at each of the clinic visits was 1.2% for both CFC- and HFA-MDIs. In conclusion, administration of a fenoterol/ipratropium bromide combination via hydrofluoroalkane-metered dose inhaler is as safe as delivery by the currently available chlorofluorocarbon-metered dose inhaler, in an extended population of patients with CAO under normal prescribing conditions.