Abstract: To address the issues of unclear or low digestion temperature, incomplete digestion and low accuracy of determination results in the current pretreatment methods for arsenic in marine organisms （national standards GB 5009.11—2014 and GB 17378.6—2007）, the method mentioned by the title was proposed. The biological sample （0.500–5.000 g） was taken and placed in a 100 mL-triangular cone bottle, and a mixture of 16 mL of nitric acid and 4 mL of perchloric acid was added. The mixture was shaken well, and then placed on an electric hot plate. The mixture was heated to 120 ℃ and digested for 0.5 h, and then heated to 200 ℃ and digested to approximately 1 mL. After cooling, 3 mL of sulfuric acid was added, and the mixture was heated to 300 ℃ and digested for more than 1 h, to approximately 0.5 mL. After cooling, the mixture was dissolved in 20 mL of 10% （volume fraction） hydrochloric acid solution at 200 ℃ to approximately 10 mL. After cooling, the digestion solution was transferred into a 25 mL-colorimetric tube, and made up its volume with water. An aliquot （5.0 mL） was placed in a 25 mL-colorimetric tube. 2.5 mL of hydrochloric acid and 2.5 mL of thiourea-ascorbic acid solution （prepared by dissolving 10 g of thiourea and 10 g of ascorbic acid in 100 mL of water） were added, settling for 15 min at room temperature, and the mixture was made its volume up with water, in which arsenic was determined by atomic fluorescence spectrometry. As shown by the results, linear relationship between the corresponding fluorescence value and mass concentration of arsenic was kept within 20 μg·L⁻¹, with detection limit （3.143s） of 0.006 mg·kg⁻¹. Six parallel analysis were conducted on the certified reference materials of marine organisms, and the determined average values of arsenic were all within the uncertainty range of the certified values, with the relative errors in the range of −2.4%−14%. Test for recovery was made on the actual samples of marine organisms, giving results in the range of 106%−108%. Six parallel analysis were conducted on the actual samples of two types of marine organisms, and RSDs of the determined values were 3.0% and 3.8% respectively. This method was compared with the second method of GB 5009.11—2014, microwave digestion-inductively coupled plasma mass spectrometry, and it was shown by the results of F-test and t-test that there was no significant difference between the two methods.