BACKGROUND The aim was to evaluate the clinical usefulness of a single plasma and bronchoalveolar lavage fluid (BALF) PCT and IL-6 measurement in discriminating septic from non-septic causes of acute respiratory distress syndrome (ARDS) and forecasting clinical outcomes. METHODS One hundred patients were enrolled within 48 h of ALI/ARDS recognition. Demographic, clinical data, severity indices were recorded and PCT and IL-6 concentrations were measured in plasma and BALF. RESULTS Plasma PCT and IL-6 values were significantly higher in septic compared to non-septic individuals (p=0.001 and 0.0005, respectively), while there were no differences in their respective BALF values. As far as identification of septic vs. non-septic ARDS is concerned, the comparison of the areas under the curves favored PCT vs. IL-6 [0.88, (95% CI 0.81-0.95) vs. 0.71, (95% CI 0.60-0.81); χ(2)=9.04, p=0.003]. A plasma PCT level of 0.815 ng/mL was associated with 74.1% sensitivity and 97.6% specificity in identifying septic ARDS cases; this corresponded to a diagnostic odds ratio value of 116. Linear regression multivariable analysis disclosed a significant relation of plasma PCT with SOFA score in septic ARDS patients (p<0.001), while neither BALF PCT nor IL-6 levels were associated with clinical outcome. CONCLUSIONS Early plasma - but not BALF - PCT concentrations can discriminate between septic and non-septic ARDS causes and are associated with the severity of multiple organ dysfunction syndrome in septic ARDS patients. However, neither plasma or BALF IL-6 levels nor BALF PCT levels carry any prognostic potential. A single plasma PCT value higher than 0.815 ng/mL makes a non-septic cause of ARDS highly unlikely.