Abstract: Palladium in palladium loaded resin was determined by inductively coupled plasma atomic emission spectrometry (ICP-AES) after the palladium loaded resins were pretreated by the ashing method, nitric acid-perchloric acid method or microwave digestion method. Three pretreatment methods were compared. The effect of sample weighing, acid amount, handling time and ashing temperature were investigated to obtain optimal experimental conditions. The interference of coexisting ions was discussed. It was found that palladium concentration in the range of 0~50 mg/L exhibited a good linearity to intensity. The linear correlation coefficient was 0.999 5. Detection limits of these three method were 0.002 8 mg/L, 0.005 4 mg/L and 0.002 3 mg/L, respectively. The relative standard deviation (RSD, n=11) were 0.69%, 0.63% and 0.42%, respectively. The recoveries were 98%~100%, 98%~101% and 100%~104%, respectively. The three methods were applied to the determination of real samples. It is found that when using the microwave digestion method the treatment time is the shortest, the sample processing is most complete and the determined values was most parallel. When both the experimental method and the atomic absorption spectrometry were applied to the same palladium loaded resin the results basically accordant with eachother.