Background: The most commonly applied treatment for acute exacerbations of chronic obstructive pulmonary disease (AECOPD) is a 5-day course of high-dose systemic corticosteroids. However, this treatment has not been shown to reduce mortality and can potentially have serious side effects. Recent research has shown that, presumably, only a subgroup of COPD patients identifieable by blood eosinophil count benefit from a rescue course of prednisolone. By applying a biomarker-guided strategy, the aim of this study is to determine whether it is possible to reduce the use of systemic corticosteroids in AECOPD without influencing the outcome. Methods: This is an ongoing prospective multicenter randomized controlled open label trial comprising 320 patients with AECOPD recruited from four hospitals in Denmark. The patients are randomized 1:1 to either standard care or eosinophil-guided corticosteroid-sparing therapy where prednisolone is not administered if the daily blood sampling reveals an eosinophil level below 0.3 × 109 cells/L. The primary endpoint is length of hospital stay within 14 days after recruitment. The secondary endpoints are treatment failure, 30-day mortality rate, COPD related re-admission rate, change in FEV1, and a number of adverse effect measures obtained within 3 months after the index hospitalisation date related to corticosteroid usage. Discussion: This will be a very large RCT providing knowledge about the effectiveness of individualized biomarker-guided corticosteroid therapy in hospitalised patients with AECOPD. Trial registration: Clinicaltrials.gov, NCT02857842 , 02-august-2016. Clinicaltrialregister.eu: Classification Code: 10,010,953, 02-marts-2016.