Abstract: A method for determination of 2-amino-4-hydroxyethylaminoanisole and its sulfate in oxidative hair dye products by high performance liquid chromatography (HPLC) was established, and high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) was used for qualitative verification. 0.5 g of hair dye cream was put into a colorimetric tube and extracted with a mixture of 2.5 g·L^-1 L-cysteine solution (used as the antioxidant) and ethanol at volume ratio of 3:7, and the volume was fixed to 10 mL with the above mixture. Waters Atlantis^® T3 MV Kit chromatographic column was used as stationary phase, and a mixture of the phosphate buffer solution containing 10% acetonitrile and acetonitrile was used as mobile phase for gradient elution. The sample solution was determined at 304 nm by diode array detector. ESI⁺ and MRM mode were used for HPLC-MS/MS analysis verification. When ion pairs of m/z 183.0>138.0 and m/z 183.0>95.2 were appeared in tested sample, and the deviation of the relative abundance ratio of ion pairs was within ±20%, 2-amino-4-hydroxyethylaminoanisole (sulfate) was confirmed in the tested sample. As shown by the results, linear relationship between mass concentration of 2-amino-4-hydroxyethylaminoanisole and its peak area was kept in the range of 1.984-496.1 mg·L^-1, with detection limits of 8 μg·g^-1 for 2-amino-4-hydroxyethylaminoanisole and 12 μg·g^-1 for 2-amino-4-hydroxyethylaminoanisole sulfate. RSDs (n=7) of peak areas of 2-amino-4-hydroxyethylaminoanisole matrix matching standard solutions with different mass concentrations in 24 h were less than 1.0%. Recovery test was made on blank matrix sample and hair dye sample by standard addition method, giving results in the range of 95.5%-101%, and RSDs (n=6) of the determined values were ranged from 1.9% to 2.4%. This method has been applied for analysis of 7 hair dye products. 2-Amino-4-hydroxyethylaminoanisoleand sulfate was detected in 5 samples, with the detection amounts in the range of 0.053 6%-0.369 0%, which was consistent with HPLC-MS/MS verification results, while it was not detected in the other 2 samples, which was inconsistent with the labeling content.