Abstract: A method of ultra-high performance liquid chromatography quadrupole electrostatic field Orbitrap mass spectrometry (UHPLC-Q/Orbitrap MS) was developed for rapid screening illegal additives in diet food. The efficacy of three extraction solvents for methanol, acetonitrile and acetonitrile containing 1% formic acid were evaluated for the extraction of illegal additives from green plums and jellies. Ultimately, acetonitrile is selected to extract illegally additives under ultrasonication. In order to achieve good separation of the additives by UHPLC in diet foods, four mobile phases were compared, including methanol (0.1% formic acid)-water (0.1% formic acid and 5 mmol/L ammonium acetate), acetonitrile (0.1% formic acid)-water (0.1% formic acid and 5 mmol/L ammonium acetate), acetonitrile-water (5 mmol/L ammonium acetate), acetonitrile-water. The effects of mobile phases and acidity levels on the chromatographic behavior of the additives were investigated, and acetonitrile-water containing 5 mmol/L ammonium acetate shows the best performance under appropriate gradient. Then, the MS data were acquired in Full MS/dd-MS² mode on UHPLC-Q/Orbitrap MS with the matrix-matched external standard method for quantification. Based on the self-constructed database of illegal additives in diet foods, one unknown additive is detected in a diet food containing green plums and jelly. By characterizing the molecular ion peak, secondary fragment ion information, and isotopic information, the unknown compound is identified to be 4-chlorobisacetophenetidin, also an illegal additive, which was verified by computation of density functional theory and analysis of the standard substance. To validate the accuracy and reliability of the method, the limits of detection (LODs), the limits of quantification (LOQs), linear ranges, recoveries, relative standard deviations (RSDs) and matrix effect (ME) for eight compounds were determined. The LODs and LOQs of deacetylbisacodyl and bisacodyl are 0.001 and 0.005 mg/kg, respectively. The LODs and LOQs of other compounds are 0.05 and 0.1 mg/kg, respectively. The correlation coefficient (R²) is higher than 0.999. All of the compounds exhibit varying degrees of MEs from moderate to weak, and the recoveries is in the range of 71.02%-103.01% with the RSDs of 0.87%-13.24%. The approach was applied to diet food containing green plums and jelly. The 4-chlorobisacodyl are detected in 30% of the green plums, suggesting that this method is suitable for the rapid qualitative and quantitative analysis of illegal additives in diet food. It is also demonstrated that the established screening process plays a positive role in identifying potential risk substances. The combination of quantum chemistry computation and mass spectrometry provides a novel strategy for the identification of unknown additives in food.