BackgroundNo evidence exists to date on which to base the selection of outcome measures for assessing nutritional interventions in critically ill patients. We conducted a systematic literature review to describe the outcomes used in recent randomised controlled trials (RCTs) assessing nutritional interventions in critically ill patients. Our objective was to set the foundation for the development of a core set of outcome measures for use in future RCTs.MethodsWe searched the PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov databases for RCTs of nutritional interventions in critically ill patients aged 18 years or older, published and/or registered between January 2000 and August 2018. Outcomes were divided into six categories (mortality, length of stay, duration of organ dysfunction, complications, functional outcomes, and others) and analysed according to the study characteristics and publication year.ResultsOf the 885 references retrieved, 170 were included in the review. Of these, 136 (80%) defined a primary outcome, 114 (67%) defined secondary outcomes (two per study on average), and 34 (20%) did not specify whether outcomes were primary or secondary. We identified 24 different outcomes in all, of which 19 were primary. Complications were the most widely used primary outcome (65/136, 48%). Mortality was the primary outcome in 17/136 (13%) studies, with six different timepoints. The main secondary outcomes were length of stay (90/114, 79%), mortality (82/114, 72%), and duration of organ dysfunction (75/114, 65%).ConclusionsThis systematic review highlights the heterogeneity of outcomes used in recent randomized controlled trials evaluating nutritional interventions in critically ill patients. The results of our systematic review may have implications for designing future RCTs of nutritional interventions in the ICU.Electronic supplementary materialThe online version of this article (10.1186/s13054-018-2303-7) contains supplementary material, which is available to authorized users.