Background: Sepsis is a condition widely observed outside critical care areas. Aim: To examine the application of sepsis screening tools for early recognition of sepsis in general hospitalized patients to: (i) identify the accuracy of these tools; (ii) determine the outcomes associated with their implementation; and (iii) describe the implementation process. Methods: A systematic review method was used. PubMed, CINAHL, Cochrane, Scopus, Web of Science, and Embase databases were systematically searched for primary articles, published from January 1990 to June 2016, that investigated screening tools or alert mechanisms for early identification of sepsis in adult general hospitalized patients. The review protocol was registered with PROSPERO (CRD42016042261). Findings: More than 8000 citations were screened for eligibility after duplicates had been removed. Six articles met the inclusion criteria testing two types of sepsis screening tools. Electronic tools can capture, recognize abnormal variables, and activate an alert in real time. However, accuracy of these tools was inconsistent across studies with only one demonstrating high specificity and sensitivity. Paper-based, nurse-led screening tools appear to be more sensitive in the identification of septic patients but were only studied in small samples and particular populations. The process of care measures appears to be enhanced; however, demonstrating improved outcomes is more challenging. Implementation details are rarely reported. Heterogeneity of studies prevented meta-analysis. Conclusion: Clinicians, researchers and health decision-makers should consider these findings and limitations when implementing screening tools, research or policy on sepsis recognition in general hospitalized patients.