Four hundred and sixty nine patients were randomized to receive either 12 micrograms bd of eformoterol (Oxis, Astra Pharmaceuticals Ltd., Kings Langley, U.K.) delivered via Turbohaler or 50 micrograms bd salmeterol (Serevent, Glaxo-Wellcome Ltd., Uxbridge, U.K.) via either the Accuhaler (Glaxo-Wellcome Ltd.) or pressurized metered dose inhaler (pMDI, Glaxo-Wellcome Ltd.) for 8 weeks. This was followed by a 4-week cross-over period when patients who had received salmeterol in the previous 8 weeks were given eformoterol and patients who had received eformoterol were given either salmeterol via the Accuhaler or pMDI to assess patient device and treatment preference. For the primary efficacy variable, the increase in peak expiratory flow (PEF) rate from run-in to 8 weeks, similar significant improvements were seen in all three treatment groups. Eformoterol Turbohaler (FT) achieved a greater increase in morning PEF than salmeterol Accuhaler (SA) from randomisation to 4 weeks; the increase shown in the eformoterol Turbohaler group was 28.9 l min-1 compared to 19.9 l min-1 for the salmeterol Accuhaler group. The addition of eformoterol Turbohaler 12 micrograms bd, to patients' existing asthma therapy was found to have a significantly more beneficial effect on the severity of patients' daytime asthma symptoms than had salmeterol Accuhaler 50 micrograms bd (P = 0.014). Eformoterol Turbohaler reduced the severity of daytime asthma symptoms by 42% after only 4 weeks of treatment. The patients in the eformoterol Turbohaler treated group experienced a higher percentage of days when they were symptom-free and did not use their short-acting bronchodilator to relieve symptoms (32.8, 24.1 and 28.0% in the FT, SA and SM groups, respectively). At 8 weeks there were no significant differences in any of these variables between the three groups. Patients in all the treatment groups gained an additional 1-1.5 nights undisturbed by asthma per week. The changes in sleep disturbance were not significantly different between the three treatment groups. In addition to the therapeutic benefits provided by eformoterol Turbohaler the device (Turbohaler) was the significant preference of patients given both Turbohaler and pMDI (P = 0.0168) and was also considered to be significantly more convenient to carry around than the Accuhaler (P < 0.0001). No other differences were found between the three devices. The results of this study demonstrate that the addition of a long-acting B2-agonist is an effective tool for achieving the goals of asthma treatment. Eformoterol via the Turbohaler is at least as effective as salmeterol via either the Accuhaler or the pMDI in achieving these goals.