Abstract: Objective To evaluate the value of central venous pressure(CVP)variability in diagnostic blood loss in critically ill patients. Methods A total of 97 critically ill patients were selected. These patients were divided into groups A, B and C according to severity grading of Acute Physiology and Chronic Health Evaluation(APACHE Ⅱ)scores. Clinical indicators were recorded, and were compared before and after blood collection. The CVP variability was detected after blood collection and its correlation with blood collection was analyzed. Results There were no significant differences in hemoglobin, mean arterial pressure and heart rate in fasting blood samples before and after first fasting blood collection(P > 0.05), while CVP showed a significant difference(P < 0.01). CVP variability was only positively correlated with blood sampling(r=0.903, P < 0.05), but showed no significant correlation with hemoglobin, mean arterial pressure, heart rate(r=0.621, P=0.320; r=0.701, P=0.231; r=0.675, P=0.256). The CVP variability and blood collection volume in group B was significantly higher than that of group A(P < 0.01), and were higher in group C was higher than that of group A and B, the differences were statistically significant(P < 0.01). Conclusion CVP variability can effectively assess the severity of diagnostic blood loss in critically ill patients. By monitoring CVP variability, measures can be taken to avoid deterioration of the disease.