Bronchiectasis guidelines recommend long-term macrolide treatment for patients with three or more exacerbations per year without Pseudomonas aeruginosa infection. Randomised controlled trials suggest that long-term macrolide treatment can prevent exacerbations in adult patients with bronchiectasis, but these individual studies have been too small to do meaningful subgroup analyses. We did a systematic review and individual patient data (IPD) meta-analysis to explore macrolide benefit in subpopulations, including those in which macrolide therapy is not currently recommended. We searched MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, and Web of Science from Jan 1, 2000, to Sept 30, 2018, to identify double blind, randomised, placebo-controlled trials of macrolide antibiotics in adult patients with bronchiectasis. We applied no language restrictions. Randomised controlled trials were eligible if treatment was defined a priori as long term and had a primary or secondary outcome of bronchiectasis exacerbations. Studies in patients with cystic fibrosis bronchiectasis were excluded. The primary outcome of the meta-analysis was frequency of exacerbations requiring treatment with antibiotics. Secondary endpoints were time to first exacerbation, change in quality of life according to the St George's Respiratory Questionnaire (SGRQ), and change in FEV1. IPD meta-analysis was done using fixed effects models adjusting for age, sex, FEV1, and trial. We did prespecified subgroup analyses for each of the primary and secondary endpoints using one-step meta-analysis only. Subgroups were defined by age, sex, previous exacerbation frequency, smoking status, inhaled corticosteroid use at baseline, body-mass index at baseline, cause, C-reactive protein at baseline, baseline FEV1 percentage of predicted, SGRQ total score, and Pseudomonas aeruginosa in sputum culture at baseline. The meta-analysis is registered with the PROSPERO international register of systematic reviews, number CRD42018102908. Of 234 identified studies, we included three randomised controlled trials, and IPD was obtained for 341 participants. Macrolide antibiotics reduced the frequency of exacerbations (adjusted incidence rate ratio [IRR] 0·49, 95% CI 0·36 to 0·66; p<0·0001). We also found that macrolide treatment improved the time to first exacerbation (adjusted hazard ratio 0·46, 0·34 to 0·61; p<0·0001) and was associated with improved quality of life measured by the SGRQ (mean improvement 2·93 points, 0·03 to 5·83; p=0·048). Macrolides were not associated with a significant improvement in FEV1 (67 mL at 1 year, -22 to 112; p=0·14). Effect estimates in prespecified subgroup analyses revealed a reduced frequency of exacerbations in all prespecified subgroups, including a high level of benefit in patients with P aeruginosa infection (IRR 0·36, 0·18-0·72; p=0·0044) and in patients with one to two exacerbations per year (0·37, 0·16-0·88; p=0·025). Studies were rated as low risk of bias across all domains. Long-term macrolide treatment significantly reduces the frequency of exacerbations in patients with bronchiectasis, with similar benefits observed in all subgroups based on patient characteristics. This finding suggests that macrolides might be considered in patients in whom macrolides are not indicated according to the current guidelines, particularly if alternative approaches to reduce exacerbations have been unsuccessful. However, downsides of long-term macrolide treatment must also be taken into account. European Respiratory Society.