The aim of the multicenter, randomized, double-blind, double-dummy, parallel-group clinical trial with a 2-week treatment period was to compare the efficacy and safety of salmeterol (50 micrograms twice daily) with slow-release (SR) terbutaline (5 mg orally, twice daily) in nocturnal asthma. A total of 159 asthmatic adults (FEV, 50-90% of predicted value; sex ratio: 0.87) with at least two nocturnal awakenings during a 7-d run-in period was included in the study. Patients were centrally randomized with a national computer network (Minitel). The main variable (number of awakening-free nights during the last week of treatment) was analyzed according to a sequential method with the one-sided triangular test. The number of awakening-free nights (+/- SD) was significantly higher in the salmeterol group: 5.3 +/- 2.4 vs 4.6 +/- 2.3 (P = 0.006). Salmeterol was significantly more effective than SR-terbutaline in the following factors: number of patients without any awakening during the last week of treatment (50% vs 27%, P = 0.003), mean morning PEF (351 +/- 109 l/min-1 vs 332 +/- 105 l/min-1, P = 0.04), PEF diurnal variation 6 +/- 10% vs 11 +/- 12%, P = 0.01), overall assessment of efficacy by the patient and the investigator (P = 0.001 and 0.005, respectively), and daily rescue salbutamol intakes (P = 0.004). In the salmeterol group, significantly fewer patients reported adverse events (16% vs 29%, P = 0.04).(ABSTRACT TRUNCATED AT 250 WORDS)