Abstract: [Objective] Aiming at the problems of complex composition of cheese sticks,cumbersome operation of traditional detection methods,and easy introduction of harmful chemicals,this study aimed to optimize the detection method of aflatoxin M₁（AFM₁） in cheese sticks. The core goals were to reduce the use of harmful chemicals,simplify operation steps,and improve detection accuracy and efficiency. [Method] Methanol was used as the extraction medium instead of traditional dichloromethane. By optimizing the pretreatment process （sample weighing,extraction,concentration and reconstitution）,combined with the Prognosis enzyme-linked immunosorbent assay （ELISA） kit for detection; the recovery rate,relative error and operation efficiency of the method were systematically evaluated by verifying spiked samples with different concentrations such as limit of quantification （LOQ）,2-fold LOQ and 5-fold LOQ. [Result] In the optimized method,the recovery rate of AFM₁ was stably between 70% and 120%,which met the requirements of GB/T 33411—2016 General Rules for Enzyme-Linked Immunosorbent Assay Kits; the relative error was controlled within ±25%,which complied with the standard of SN/T 2775—2023 Evaluation Method for Commercial Food Detection Kits.The pretreatment time was shortened from 70 min of the traditional method to 25 min,and the operation steps were simplified from 4 steps to 2 steps. [Conclusion] The optimized methanol extraction-ELISA detection method has good extraction effect,detection accuracy and efficiency,can meet the batch detection needs of dairy enterprises,and provides accurate and reliable technical support for the detection of AFM₁ content in cheese sticks.

Key words: methanol extraction, aflatoxin M₁（AFM₁）, cheese stick, enzyme-linked immunosorbent assay, method optimization