INTRODUCTION: Enteral nutrition (EN) with eicosapentaenoic acid (EPA)/γ-linolenic acid (GLA) is recommended for mechanically ventilated patients with severe lung injury. EPA/GLA has anti-inflammatory benefits, as evidenced by its association with reduction in pulmonary inflammation, improvement in oxygenation and improved clinical outcomes in patients with severe forms of acute lung injury. This study was a prospective, multicenter, randomized, double-blinded, controlled trial designed to investigate whether EPA/GLA could have an effective role in the treatment of patients with early sepsis (systemic inflammatory response syndrome with confirmed or presumed infection and without any organ dysfunction) by reducing the progression of the disease to severe sepsis (sepsis associated with at least one organ failure) or septic shock (sepsis associated with hypotension despite adequate fluid resuscitation). Secondary outcomes included the development of individual organ failure, increased ICU and hospital length of stay, need for mechanical ventilation and 28-day all-cause mortality. METHODS: Randomization was concealed, and patients were allocated to receive, for seven days, either an EPA/GLA diet or an isocaloric, isonitrogenous control diet not enhanced with lipids. Patients were continuously tube-fed at a minimum of 75% of basal energy expenditure × 1.3. To evaluate the progression to severe sepsis and/or septic shock, daily screening for individual organ failure was performed. All clinical outcomes were recorded during a 28-day follow-up period. RESULTS: A total of 115 patients in the early stages of sepsis requiring EN were included, among whom 106 were considered evaluable. Intention-to-treat (ITT) analysis demonstrated that patients fed the EPA/GLA diet developed less severe sepsis and/or septic shock than patients fed the control diet (26.3% versus 50%, respectively; P = 0.0259), with similar results observed for the evaluable patients (26.4% versus 50.9% respectively; P = 0.0217). The ITT analysis demonstrated that patients in the study group developed cardiovascular failure (36.2% versus 21%, respectively; P = 0.0381) and respiratory failure (39.6% versus 24.6%, respectively; P = 0.0362) less often than the control group. Similarly, when considering only the evaluable patients, fewer patients developed cardiovascular failure (20.7% versus 37.7%, respectively; P = 0.03) and respiratory failure (26.4% versus 39.6%, respectively; P = 0.04). The percentage of patients fed the EPA/GLA diet requiring invasive mechanical ventilation was reduced compared with controls (ITT patients: 18.9% versus 33.9%, respectively; P = 0.394; evaluable patients: 17.5% versus 34.5%, respectively; P = 0.295). Patients nourished with the EPA/GLA diet remained in the ICU fewer days than the control population (ITT patients: 21.1 ICU-free days versus 14.7 ICU-free days, respectively; P < 0.0001; evaluable patients: 20.8 ICU-free days versus 14.3 ICU-free days, respectively; P < 0.0001) and fewer days at the hospital (ITT patients: 19.5 hospital-free days versus 10.3 hospital-free days, respectively; P < 0.0001; evaluable patients: 19.1 hospital-free days versus 10.2 hospital-free days, respectively; P < 0.001) (all numbers expressed as means). No significant differences in 28-day all-cause mortality were observed (ITT patients: 26.2% EPA/GLA diet versus 27.6% control diet, respectively; P = 0.72; evaluable: 26.4 EPA/GLA diet versus 30.18 control diet, respectively; P = 0.79). CONCLUSIONS: These data suggest that EPA/GLA may play a beneficial role in the treatment of enterally fed patients in the early stages of sepsis without associated organ dysfunction by contributing to slowing the progression of sepsis-related organ dysfunction, especially with regard to cardiovascular and respiratory dysfunction. TRIAL REGISTRATION: ClinicalTrials.gov: NCT00981877.