Background In February, 2017, the US Food and Drug Administration approved the blood infection marker procalcitonin for guiding antibiotic therapy in patients with acute respiratory infections. This meta-analysis of patient data from 26 randomised controlled trials was designed to assess safety of procalcitonin-guided treatment in patients with acute respiratory infections from different clinical settings. Methods Based on a prespecified Cochrane protocol, we did a systematic literature search on the Cochrane Central Register of Controlled Trials, MEDLINE, and Embase, and pooled individual patient data from trials in which patients with respiratory infections were randomly assigned to receive antibiotics based on procalcitonin concentrations (procalcitonin-guided group) or control. The coprimary endpoints were 30-day mortality and setting-specific treatment failure. Secondary endpoints were antibiotic use, length of stay, and antibiotic side-effects. Findings We identified 990 records from the literature search, of which 71 articles were assessed for eligibility after exclusion of 919 records. We collected data on 6708 patients from 26 eligible trials in 12 countries. Mortality at 30 days was significantly lower in procalcitonin-guided patients than in control patients (286 [9%] deaths in 3336 procalcitonin-guided patients vs 336 [10%] in 3372 controls; adjusted odds ratio [OR] 0·83 [95% CI 0·70 to 0·99], p=0·037). This mortality benefit was similar across subgroups by setting and type of infection (p interactions >0·05), although mortality was very low in primary care and in patients with acute bronchitis. Procalcitonin guidance was also associated with a 2·4-day reduction in antibiotic exposure (5·7 vs 8·1 days [95% CI −2·71 to −2·15], p<0·0001) and a reduction in antibiotic-related side-effects (16% vs 22%, adjusted OR 0·68 [95% CI 0·57 to 0·82], p<0·0001). Interpretation Use of procalcitonin to guide antibiotic treatment in patients with acute respiratory infections reduces antibiotic exposure and side-effects, and improves survival. Widespread implementation of procalcitonin protocols in patients with acute respiratory infections thus has the potential to improve antibiotic management with positive effects on clinical outcomes and on the current threat of increasing antibiotic multiresistance. Funding National Institute for Health Research.