The aim of the study was to investigate whether procalcitonin, soluble CD14 and interleukin-6 show advantages in predicting the outcome and specificity for bacterial infection in patients with sepsis in comparison to common C-reactive protein measurement. Laboratory parameters were measured in plasma of patients during 14 days following the diagnosis of sepsis. Patients fulfilling the ACCP/SCCM criteria for sepsis were admitted to an intensive care unit (n=35). Procalcitonin was measured with an immunoluminometric assay, and soluble CD14 and interleukin-6 were analysed by ELISA. C-reactive protein was determined nephelometrically. Measurements were performed on days 0, 1, 2, 3, 4, 7 and 14. Separating the patients into survivors (n=22) and non-survivors (n=13), it was demonstrated that non-survivors mostly exhibited, after the day of admission, increasing procalcitonin concentrations which peaked around days three and four. In contrast, the procalcitonin concentrations of survivors fell continuously to the value of 2.1 ng/ml which was reported to be important for patients prognosis. The difference between procalcitonin median values of survivors (n=22) and non-survivors (n=13) attained the level of statistical significance on day 7 and on day 14 (p=0.05). When comparing the median values of C-reactive protein, soluble CD14 and interleukin-6 between survivors and non-survivors, no significant differences were detectable. In this study, plasma concentrations of soluble CD14 and interleukin-6 showed no predictive value for patients' outcome as compared with established laboratory parameters such as C-reactive protein or leukocyte count. Monitoring of procalcitonin seemed to detect severe episodes of sepsis and may improve the laboratory monitoring of septic patients.