Abstract: Honey has been used as a food and medicine for millennia. The free aromatic amino groups of sulfonamides can easily react with the main components of honey (glucose and fructose) to form monosaccharide-bound sulfonamides, which directly affect the quantitative results of sulfonamide residues in honey. This presence of bound residues should be also a matter of concern for public health since the free drugs can be liberated by acid and enzymatic hydrolysis during the digestion in stomach. Thus, the accurate determination of masked sulfonamides content in honey is essential to comply with official residue control. Acid hydrolysis is currently the most popular approach to restore the bound sulfonamides to their free forms. And spiking recovery was widely used for the optimization of hydrolysis step. However, if binding reaction couldn′t occur immediately and even not occurred after standard analytes spiking into blank matrix in vitro, the spiked matrix could have specific distinctions with real contaminated matrix in which the binding reaction had already reached dynamic balance. In this paper, recovery in spiked honey and quantitative result in positive honey were taken as the investigation index of hydrolysis step at the same time. To improve the accuracy of quantitative results of sulfonamide residues in honey, three typical sample preparation methods containing different hydrolysis steps were compared via ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS). The results showed that sulfonamides could not be fully recovered from spiked honey whatever during the extraction with 0.7 mol/L HCl or 0.3 mol/L citric acid in 45 ℃ water bath for 1 h followed by Oasis Prime HLB cleanup or during the extraction phosphoric acid solution (pH 2) followed by Oasis MCX and Oasis HLB column double purification. The absolute recoveries were 59.2%-72.2%, 27.8%-71.7% and 0.4%-48.9%, respectively. As matrix effect could lead to lower measurements, isotope dilution mass spectrometry (IDMS) were used to simultaneously correct the loss of sulfonamides and the interference of matrix effect. The relative recoveries of three sample preparation methods were between 95.4% and 104.2%, the correction factor of IDMS is 0.95-1.03. Hydrolysis extraction with 0.7 mol/L hydrochloric acid in a water bath for 1 h followed by Oasis Prime HLB cleanup provided the best quantitative results of 5 sulfonamides in positive honey and therefore proved to be the optimal sample preparation method. But when 0.3 mol/L citric acid or phosphoric acid solution was used instead for hydrolysis, even though the relative recovery was higher than 95.4%, the quantitative results in positive honey were still significantly lower by 23.1%-57.3% and 73.8%-99.6%, respectively. It surely means the sugar bound sulfonamides cannot be completely liberated during the hydrolysis process. This implies the hydrolysis efficiency obtained by the use of spiking recovery have less validity. Thus, positive samples are essential for the verification of hydrolysis method.