The aim of this prospective study was to investigate whether serum procalcitonin (PCT) can be used as a post-mortem marker of sepsis and to determine whether this biochemical parameter can be employed in the forensic elucidation of death due to sepsis. At least three blood samples were collected between 0.3 and 139 h post-mortem from sepsis-related fatalities (n = 8) and control individuals (n = 53, where death was due to various natural and unnatural causes). Additionally one ante-mortem blood sample was collected shortly before death from the patients in the sepsis group. In the sepsis group, serum PCT concentrations, determined by using an immunoluminometric assay, were elevated in all patients for the whole observation period, whereas in the control group serum PCT was not detectable in 94% of the cases. Measurement of PCT levels seems reasonable until at least approximately 140 h postmortem, depending on the ante-mortem levels. A linear regression model is presented that allows the serum PCT concentration of an individual at the time of death to be estimated on condition that at least two positive post-mortem PCT values have been determined. Ante-mortem PCT values correlated well with the predicted PCT values at the time of death in the sepsis group using the standardized PCT logarithms. According to the results of the present study, PCT is a valuable biochemical parameter for the post-mortem discrimination between sepsis and underlying non-septic causes of death.