Management of severe sepsis, an acute illness with high morbidity and mortality, suffers from the lack of effective biomarkers and largely empirical predictions of disease progression and therapeutic responses. We conducted a genome-wide association study using a large randomized clinical trial cohort to discover genetic biomarkers of response to therapy and prognosis utilizing novel approaches, including combination markers, to overcome limitations of single-marker analyses. Sepsis prognostic models were dominated by clinical variables with genetic markers less informative. In contrast, evidence for gene-gene interactions were identified for sepsis treatment responses with genetic biomarkers dominating models for predicting therapeutic responses, yielding candidates for replication in other cohorts.