Objective To evaluate the clinical value of serum procalcitonin (PCT) in the diagnosis of bacte-rial infections in the critically ill patients. Methods A loud d 100 cases of critical patients were divided into bacteri-al infection group(56 cases) and non-bacterial infection group(44 cases). Serum PCT was measured by immunolu-minometric assay. Results The concentration of PCT in bacterial infection group(21.54 ±5.72) μg/L, was signifi-cantly higher than non-bacterial infection group (11. 95±4. 58)μg/L (t =2.291,P<0.05);The APACHE Ⅲ score of bacterial infection group(62. 44 ± 19. 55) cent was significantly higher than non-bacterial infection group(44. 56 ± 25. 88) cent(t = 2. 195 ,P < 0. 05). The concentration of PCT of 1.0μg/L and 2. 0 p.g/L compared to the former sensitivity (96. 5) % higher than the latter (55.2) % (X2 = 3. 94, P < 0. 05), the former specificity (41.7) % lower than the latter (95.8)% (X2 = 4. 02 ,P < 0. 05);1.5μg/L as a positive standard Youden index and the Agreement (83.0 % ,0. 65), were significantly higher than 1 μg/L and 2 μg/L(71.7%, 0. 38 ;73.6% ,0. 51) (X2 = 3.84, X2=3. 90,X2 = 3.992 = 3.91 ,P < 0. 05);The concentration of PCT in death group (38. 9 ±12. 6)μg/L was signifi-canfly higher than the survival group(11.8± 8. 3) μg/L(t =2. 398 ,P <0. 05). Conclusion Serum PCT has clinical values in the diagnosis and therapy of bacterial infections in the critical patients.