【Objective】 To establish and optimize a method for the determination of aluminum (Al) residue by inductively coupled plasma mass spectrometry (ICP-MS). 【Methods】 Nitric acid solution was used to treat samples and standards. The concentration of nitric acid solution and equipment parameters were optimized, and the specificity, linearity, repeatability, accuracy, detection limit, quantitative limit and intermediate precision of the optimized detection method were investigated to confirm whether it was suitable for the determination of Al residue in human serum albumin. 【Results】 The concentration of nitric acid was 5%, and digest time was 4 h. The equipment condition of ICP-MS was as follows: RF power: 1600 W, sampling depth: 10 mm, atomizer / carrier gas flow rate: 1.0 L/ min, compensation flow rate: 0.5 L/ min, experimental mode: standard mode, integration time: 0.2 s, data acquisition: 3 times. Specificity: The recoveries of Al: 92% (high concentration, RSD=3.5%), 98% (low concentration, RSD=4.9%). Linearity: In the range of (0~40) μg/L, the correlation coefficient between concentration and optical energy signal (CPS) of standard / sample were higher than 0.999 0. Accuracy/ Repeatability: The recoveries of sample (3 concentration): 108% (RSD=4.7%), 110% (RSD=4.9%) and 110% (RSD=2.8%). The detection limit was 0.006 μg/L, and the quantitation limit was 0.019 μg/L. Intermediate precision: personnel factor and date factor, P>0.05, RSD (12 times)=2%. Comparison between ICP-MS and atomic absorption spectrometry (AAS): the deviation between ICP-MS and AAS was 8%, and that of samples was 3%, with no significant difference noticed between the two methods. 【Conclusion】 After optimization, ICP-MS method has shown good performance in terms of specificity, linearity, repeatability, accuracy, detection limit, quantitative limit and intermediate precision, and is suitable for the determination of Al residue in human albumin products of our company.