Background: Sepsis-associated encephalopathy (SAE) occurs in sepsis survivors, where neuronal inflammation plays a critical role in its pathogenesis. Our recent studies demonstrated that plasma extracellular (ex) miRNAs, including miR-146a, miR-122, miR-34a, miR-145, increased in sepsis. These ex-miRNAs cause neuronal inflammation both in vitro in the microglia and in vivo in the intact brain. Extracellular vehicles (EVs) have been proposed as a vehicle for ex-miRNA-mediated intercellular communication. However, the biological function of plasma EVs and their miRNA cargo in neuronal inflammation is unclear. In the current study, we hypothesize that septic plasma EVmiRNAs contribute to brain inflammation. Methods: Sepsis was created by cecal ligation and puncture (CLP) surgery in wild-type (WT) mice. EVs were isolated from plasma at 24 hours after CLP using ultracentrifugation and quantified using a nanotracker, Viewsizer 3000 (Horiba). Cell infiltration and cytokine gene expression were tested by flow cytometry and qRT-PCR following intracerebroventricular injection (ICV) of plasma EVs (8.2x10^8 particles/brain). Microglia were isolated from neonatal brain of WT mice and treated with EVs from sham or CLP plasma at various doses. The function of EVs associated miRNAs was tested by preincubating EVs with control oligos or anti-miRNAs combo (anti-miR-146a, -122, -34a, -145) prior to the cell treatment. Results: ICV delivery of septic plasma EVs led to a marked increase in immune cell infiltration to the mouse brain including neutrophils and monocytes both in the numbers and percentage as compared to that of sham plasma EVs (sham vs. sepsis EVs: Neutrophil: 9.5 ± 2.3 vs. 20.1 ± 4.5 x 10^3, monocyte: 7.8 ± 2.6 vs. 15 ± 3 x 10^4, p<0.01). Cytokine/chemokine gene expressions, such as CXCL2, TNFα, IL-6, IL-1 β, were also significantly upregulated about 1.5-2-fold in the cerebral cortex following ICV injection of sepsis plasma EVs. In vitro, EVs isolated from CLP plasma induced a robust cytokine production including CXCL2 and IL-6 in a dose-dependent manner in the microglia, while sham plasma EVs did not possess any proinflammatory effect. Pretreatment of EVs with anti-miRNAs combo (antimiR-146a, -122, -34a, -145) at dose of 200 nM led to a 49.3% inhibition in septic plasma EV-mediated CXCL2 production as compared to that of control oligos. Conclusion: In the current study, we demonstrated that plasma EVs of septic mice cause brain inflammation both in the intact animal and in the microglial cultures. The proinflammatory property of sepsis plasma EVs is in part attributed to their cargo miRNAs. Future studies will investigate how circulating EVs cross blood brain barrier and the mechanisms by which EV miRNA cargo contributes to neuronal inflammation during sepsis.