Background: Treatment guidelines for patients with moderate persistent asthma recommend regular therapy with a combination of an inhaled corticosteroid and a longacting β2 agonist plus as-needed rapid-acting bronchodilators. We investigated whether symptom-driven budesonide and formoterol combination therapy administered as needed would be as effective as regular treatment with this combination plus as-needed symptom-driven terbutaline for patients with moderate asthma. Methods: In this non-inferiority randomised clinical trial, we recruited adult patients (18-65 years of age) with stable moderate persistent asthma, according to 2006 Global Initiative for Asthma guidelines. Patients were recruited from outpatient clinics of secondary and tertiary referral hospitals and university centres. After a 6-week run-in period of inhaled regular budesonide and formoterol plus as-needed terbutaline, the patients were randomly assigned in a 1:1 ratio to receive placebo twice daily plus as-needed treatment with inhaled 160 μg budesonide and 4·5 μg formoterol (as-needed budesonide and formoterol therapy) or twice-daily 160 μg budesonide and 4·5 μg formoterol combination plus symptom-driven 500 μg terbutaline (regular budesonide/formoterol therapy) for 1 year. Randomisation was done according to a list prepared with the use of a random number generator and a balanced-block design stratified by centre. Patients and investigators were masked to treatment assignment. The primary outcome was time to first treatment failure measured after 1 year of treatment using Kaplan-Meier estimates, and the power of the study was calculated based on the rate of treatment failure. Analyses were done on the intention-to-treat population. This study is registered with ClinicalTrials.gov, number NCT00849095. Findings: Between April 20, 2009, and March 31, 2012, we screened 1010 patients with moderate asthma and randomly assigned 866 eligible patients to the two treatment groups (424 to as-needed budesonide and formoterol therapy and 442 to regular budenoside and formoterol therapy). Compared with regular budesonide and formoterol therapy, as-needed budesonide and formoterol treatment was associated with a lower probability of patients having no treatment failure at 1 year (Kaplan-Meier estimates 53·6% for as-needed treatment vs 64·0% for regular treatment; difference 10·3% [95% CI 3·2-17·4], at a predefined non-inferiority limit of 9%). Patients in the as-needed budesonide and formoterol group had shorter time to first treatment failure than those in the regular therapy group (11·86 weeks vs 28·00 weeks for the first quartile [ie, the time until the first 25% of patients experienced treatment failure]). The difference in treatment failures was largely attributable to nocturnal awakenings (82 patients in the as-needed treatment group vs 44 in the regular treatment group). Both treatment regimens were well tolerated. Interpretation: In patients with moderate stable asthma, as-needed budesonide and formoterol therapy is less effective than is the guideline-recommended regular budesonide and formoterol treatment, even though the differences are small. Funding: Italian Medicines Agency.