The use of procalcitonin (PCT) as a tool to assist clinicians in using antibiotics in intensive care patients has been postulated. Here we evaluate the efficacy of procalcitonin determination in helping clinicians in the decision to start or discontinue an antibiotic treatment in patients admitted to infectious disease wards. A retrospective observational single centre study was conducted in two infectious disease wards. Descriptive and inferential statistical analysis was carried out and receiver operating characteristic curves and area under the curve (AUC) were used to assess the accuracy of PCT and C-reactive protein (CRP) in separating patients undergoing antibiotic treatment or otherwise. In all, 164 patients were analysed of whom 99 (60.4%) were not on antibiotic treatment at the time of PCT determination, whereas 65 (39.6%) took antibiotics. Regarding the accuracy of PCT and CRP in determining a subsequent antibiotic prescription in patients without an ongoing antibiotic treatment, no statistically significant difference between the two markers was detected [AUC, 0.75; confidence interval (CI) 95%: 0.66-0.84; vs 0.69; CI 95%: 0.59-0.79 for PCT and CRP, respectively; p=0.32]. Conversely, in patients with an ongoing antibiotic treatment a statistically significant difference between PCT and CRP AUC in their ability to determine an antibiotic interruption was observed [0.77 (CI 95%: 0.65-0.89) vs 0.59 (CI 95%: 0.45-0.73) (p=0.03)]. PCT determination appeared to be more helpful than CRP in determining discontinuation of an antibiotic treatment in non-intensive care patients. However, PCT should supplement and not supplant a careful clinical evaluation.