BACKGROUND: Modest benefits of antibiotics for acute upper respiratory tract infections have to be weighed against common adverse reactions, cost and antibacterial resistance. There has been interest in ways to reduce antibiotic prescribing. One strategy is to provide the prescription, but advise delay of more than 48 hours before use, in the hope symptoms resolve first. Advocates suggest this will preserve patient satisfaction. This review asks what effect delayed antibiotics have on clinical outcomes of respiratory infections, antibiotic use and patient satisfaction. OBJECTIVES: To evaluate the prescribing strategy of delayed antibiotics for acute respiratory tract infections compared to immediate or no antibiotics for clinical outcomes, antibiotic use and patient satisfaction. SEARCH STRATEGY: We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, Issue 4, 2006); MEDLINE (January 1966 to January Week 2, 2007), EMBASE (1990 to Week 2, 2007) and Current Contents - ISI Web of Knowledge (1998 to January 2007). SELECTION CRITERIA: Randomised controlled trials (RCTs) involving patients of all ages defined as having an acute respiratory infection were included in which delayed antibiotics were compared to antibiotics used immediately or no antibiotics. Outcomes measured included clinical outcomes, antibiotic use and patient satisfaction. DATA COLLECTION AND ANALYSIS: Data were collected and analysed by three review authors. MAIN RESULTS: Nine trials were eligible on the basis of design and relevant outcomes. For most clinical outcomes there was no difference between delayed, immediate and no antibiotics. Antibiotics prescribed immediately were more effective than delayed for fever, pain and malaise in some studies of patients with acute otitis media and sore throat but for other studies there was no difference. There was no difference for the common cold and bronchitis. Delaying antibiotic prescriptions reduced antibiotic use, and in three studies, reduced patient satisfaction compared to immediate antibiotics. In the other two studies comparing delayed and immediate antibiotics measuring satisfaction, there was no difference. Two studies also included a 'no antibiotics' arm for bronchitis and sore throat: there was no difference in symptom resolution nor patient satisfaction from antibiotic delay. In one study, but not the other, antibiotic use was significantly decreased with no, rather than delayed, antibiotics. AUTHORS' CONCLUSIONS: For most clinical outcomes there is no difference between the strategies. Immediate antibiotics was the strategy most likely to provide the best clinical outcomes in patients with sore throat and otitis media. Delaying or avoiding antibiotics, rather than providing them immediately, reduces antibiotic use for acute respiratory infections. Delay also reduced patient satisfaction in three trials, compared to immediate antibiotics with no difference in two other trials. Delaying antibiotics seems to have little advantage over avoiding them altogether where it is safe to do so.