BACKGROUND Procalcitonin (PCT) is an emerging prognostic marker in coronavirus disease 2019 (COVID-19). Whether PCT can detect secondary bacterial infections or reflect target tissue injury in this setting is still unclear. Here we performed a meta-analysis to review the prognostic value of PCT for severe disease and adverse outcome events in COVID-19. METHODS We searched relevant publications in online databases. Studies were included if they reported categorical data according to disease severity and/or outcomes. We analysed extracted data using fixed or random-effects meta-analysis models, as appropriate, depending on the presence of significant heterogeneity. RESULTS Data from 14 studies (3492 patients) were included in the analysis. Overall, 163 of 256 patients with elevated PCT had severe disease (63.7%) compared with 553 of 2047 with negative PCT (27.0%) (OR: 5.92; 95% CI: 3.20 to 10.94). Elevated PCT was also associated with adverse outcomes (OR: 13.1; 95% CI: 7.37 to 23.1). PCT was increased in 22.8% and 30.6% of patients with the severe course and adverse outcome, respectively. Rates of secondary bacterial infections ranged from 4.7% to 19.5% and were associated with increased risk of severe course or fatal outcomes (OR: 20.8; 95% CI: 11.6 to 37.4). CONCLUSIONS Elevated PCT levels could identify a subset of COVID-19 patients at increased risk of severe disease and adverse outcome. Its limitations include low sensitivity and undefined cost-utility ratio. Whether PCT may be used for detecting secondary bacterial infections and guiding antibiotic therapy in COVID-19 is still undefined.