Sepsis is common, deadly, and costly. Over 1 million patients are affected each year, and as many as half of them die. The cost of care exceeds that of any hospital admission. Early diagnosis and rapid response are essential elements of effective treatment. Nurses providing direct patient care have the patient contact and clinical knowledge to make them critical components of inpatient sepsis prevention, surveillance, and response. There is a large research literature on sepsis. Many studies evaluate clinical interventions and examine patient risk factors. These studies inform evidence-based guidelines, such as the Surviving Sepsis Campaign. Despite the international expert consensus on sepsis treatment that this campaign represents, sepsis incidence and mortality varies by hospital. The Quality Health Outcomes Model posits that system (and patient) characteristics mediate the relationship between interventions and outcomes such that in actual practice, clinical guidelines often do not have their intended effects. Hospital nursing characteristics are system-level features that may help explain institutional differences in sepsis incidence and mortality. This study explored the relationship between hospital nursing characteristics and sepsis. Specifically, it determined the impact of nurse staffing, education, and the work environment among postsurgical patients on the odds of sepsis (Aim 1), and on mortality among septic patients (Aim 2). This was a cross-sectional, secondary analysis of nurse survey responses, patient discharge abstracts, and hospital administrative data from hospitals in four states. The sample included 1,435,919 patients who were hospitalized for general, orthopedic, or vascular surgeries from 2005 through 2007, 23,603 nurse survey respondents, and the 503 hospitals associated with these subjects. Logistic regression was used to model the relationship between hospital nursing characteristics and patient outcomes. There was a significant association between hospital work environment and postsurgical sepsis and between nurse education and death after sepsis. Surgical patients in hospitals with better nurse work environment experienced lower odds of sepsis (OR 0.93; p=0.002). Postsurgical septic patients in hospitals with a higher percentage of BSN-prepared nurses had lower odds of death (OR 0.94; p<0.001). Nursing resources were associated with patient outcomes and may be a mechanism for administrators to reduce sepsis incidence and mortality.