This study measured Procalcitonin (PCT), Presepsin (PRE-S) and pro-Adrenomedullin (pro-ADM) in intensive care unit (ICU) patients blood to assess their contribution to accurate diagnosis of sepsis and potential predictive impact on prognosis. The final aim was to improve the use of infection biomarkers for optimizing the impact of laboratory medicine on clinical outcomes, focusing on the good management of resources designed to produce maximum effectiveness and efficiency. Sixty-four adult patients were studied during their hospitalization in ICU; blood samples were collected and categorized according to their clinical diagnosis and illness severity, and sepsis marker levels were measured on automated immunoassay platforms. PCT, PRE-S and pro-ADM infection markers were significantly lower in controls than in sepsis or septic shock groups. The area under the curve, by ROC curve analysis, was 0.945 for PCT, 0.756 for PRE-S and 0.741 for pro-ADM. Sepsis diagnostic accuracy was not improved by combining PCT, PRE-S and pro-ADM measures. Preliminary data demonstrated that, despite PRE-S and pro-ADM being able to differentiate between septic and non-septic patients with accuracy, PCT remains the most reliable marker available. The results obtained still do not allow us to consider a combination of markers, because it would merely increase laboratory costs without improving diagnostic performance. Furthermore, the results confirm a possible prognostic role of pro-ADM in septic states, but no correlation between biomarker levels and survival at 48 h was detected. Hence PCT, PRE-S, nor pro-ADM can be used to predict short-term prognosis.