Abstract: Objective To find better indicator of pre-stage renal injury in cadmium exposed population around a factory in Huangshi city.Methods Blood and urine samples from 649 participators were collected,and areas under the curves of blood and urine cadmium were calculated with SPSS softwaer.Then with standardized partial regression,coefficients of logistic regression model were used for judement of the indicators.Results Among 649 samples,area under the receiver operating characteristic curves of blood cadmium(0.569)was not as large as that of urinary cadmium(0.725).And urine cadmium!s standardized partial regression coefficient(0.5036)was 23.75 times higher that that of blood cadmium's (-0.0212)in logistic regression model.Conclusion Urine cadmiumis superior to blood cadmium in reflecting cadmiuminduced renal dysfunction.