Sepsis is primarily a state of uncontrolled systemic inflammation. Despite decades of sepsis research involving mammalian models and over 100 clinical trials, sepsis remains a major healthcare burden. Severe sepsis strikes the young and old alike with an incidence of ~750,000 cases per year worldwide, and an estimated mortality rate of 50-80%. As an alternative to traditional animal models, zebrafish have recently emerged as a powerful vertebrate paradigm to study human pathologies and for phenotype-based high-throughput drug screening. Here we show that sepsis can be effectively modeled in the zebrafish. Lipopolysaccharide (LPS) was used to induce sepsis-like pathology in 3dpf zebrafish by water delivery. Through the use of several tissue-specific fluorescent-reporter mediated transgenic lines, fluorescent microangiography, histological assessment, and gene expression analyses, we strived to assess LPS-induced systemic inflammation. LPS delivery to zebrafish embryos at a concentration of 100µg/ml effectively models the dynamics of sepsis progression, leading to 90% mortality. Our zebrafish sepsis model exhibits the major hallmarks of human sepsis including widespread vascular leakage, edema and tissue damage, increased inflammatory cell infiltration and reactive oxygen species (ROS) production, reduced blood circulation and increased thrombocyte aggregation. We validated the suitability of the model for phenotype-based drug screening using fasudil, a drug known to rescue LPS induced vascular permeability in non-primate sepsis models. LPS induced mortality, tissue edema and ROS production were identified as fast and reliable read-outs for high-throughput drug screening. This novel zebrafish sepsis model is expected to provide unmatched potential to screen and validate large numbers of compounds that can modify sepsis pathology in vivo.