Sepsis represents a major cause of lethality during intensive care unit (ICU) treatment. Pharmacological treatment strategies for sepsis are still limited and mainly based on the early initiation of antibiotic and supportive treatment. In this context, numerous clinical and serum based markers have been evaluated for the diagnosis, the severity, and the etiology of sepsis. However until now, few of these factors could be translated into clinical use. MicroRNAs (miRNAs) do not encode for proteins but regulate gene expression by inhibiting the translation or transcription of their target mRNAs. Recently it was demonstrated that miRNAs are released into the circulation and that the spectrum of circulating miRNAs might be altered during various pathologic conditions, such as inflammation, infection, and sepsis. By using array- and single PCR-based methods, a variety of deregulated miRNAs, including miR-25, miR-133a, miR-146, miR-150, and miR-223, were described in the context of sepsis. Some of the miRNAs correlated with the disease stage, as well as patients' short and long term prognosis. Here, we summarize the current findings on the role of circulating miRNAs in the diagnosis and staging of sepsis in critically ill patients. We compare data from patients with findings from animal models and, finally, highlight the challenges and drawbacks that currently prevent the use of circulating miRNAs as biomarkers in clinical routine.