Objective The aim of this study was to establish a scientific and reasonable estimation and allocation model for food safety evaluation.Methods A statistical method was used to estimate the total number of samples for evaluative sampling， a sample size allocation index system was established based on an analysis of the literature， the Delphi method was used to screen multiple factors， and a combination weight was assigned to each index based on the analytic hierarchy process and CRITIC method.Results Based on the estimated failure rate of food safety evaluation sampling in Province X， the total number of food safety evaluation samples in Province X was determined to be 7 800. The evaluation sampling size allocation index system covered two aspects of food varieties and testing areas and included the following eight indicators： sampling failure rate， dietary structure， the number of food safety incidents， attention， food consumption， consumption expenditure， and resident population. The allocation of samples for each food category in each prefecture-level city in the evaluative sampling test in Province X was obtained based on the weights of the calculated indicator combinations.Conclusion The sample size allocation model can be used to allocate the sample size in a more objective and reasonable manner according to the food safety situation in each region and can be used as a reference for food safety evaluation sampling.