The objective of the study was to determine the accuracy of phospholipase A2 group II (PLA2-II), interferon-gamma-inducible protein 10 (IP-10), angiopoietin-2 (Ang-2), and procalcitonin (PCT) plasma levels in early ruling in/out of sepsis among systemic inflammatory response syndrome (SIRS) patients. Biomarker levels were determined in 80 SIRS patients during the first 4 h of admission to the medical ward. The final diagnosis of sepsis or non-infective SIRS was issued according to good clinical practice. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for sepsis diagnosis were assessed. The optimal biomarker combinations with clinical variables were investigated by logistic regression and decision tree (CART). PLA2-II, IP-10 and PCT, but not Ang-2, were significantly higher in septic (n = 60) than in non-infective SIRS (n = 20) patients (P ≤ 0.001, 0.027, and 0.002, respectively). PLA2-II PPV and NPV were 88 and 86%, respectively. The corresponding figures were 100 and 31% for IP-10, and 93 and 35% for PCT. Binary logistic regression model had 100% PPV and NPV, while manual and software-generated CART reached an overall accuracy of 95 and 98%, respectively, both with 100% NPV. PLA2-II and IP-10 associated with clinical variables in regression or decision tree heterogeneous models may be valuable biomarkers for sepsis diagnosis in SIRS patients admitted to medical ward (MW). Further studies are needed to introduce them into clinical practice.