BACKGROUND : Fluid resuscitation is a widely used intervention that is mandated in the management of sepsis. While its use can be life-saving, its overuse is associated with harm. Despite this, the best means of assessing a need for fluid resuscitation in an acute medical setting is unclear. OBJECTIVE : To assess studies of diagnostic tests that identify the need for fluid resuscitation in adults with sepsis, as defined by the presence of fluid responsiveness. METHODS : Protocol registration was performed in advance (PROSPERO:CRD42017048651). Research database searches were performed alongside additional searches to identify grey literature. Diagnostic test accuracy studies that assessed any fluid assessment tool were identified independently by two authors, before data extraction and quality assessments were performed. METHODS : Adults with sepsis, without intensive care organ support, who would be appropriate for admission to an acute medical unit. RESULTS : Of the 26,841 articles that were screened, 14 studies were identified for inclusion, involving a combined total of 594 patients. Five categories of index test were identified: inferior vena cava collapsibility index (IVCCI), haemodynamic change with passive leg raise, haemodynamic change with respiration, haemodynamic change with intravenous fluid administration, and static assessment tools. Due to the high level of clinical heterogeneity affecting all aspects of study design, quantitative analysis was not feasible. There was a lack of consensus on reference tests to determine fluid responsiveness. CONCLUSIONS : While fluid resuscitation is considered a key part of the management of sepsis, evidence to support fluid assessment in awake adults is lacking. This review has highlighted a number of research recommendations that should be addressed as a matter of urgency if patient harm is to be avoided.