Objective To analyze the correlation of muti-mycotoxin contamination in wheat， the co-contamination characteristics of different mycotoxins were studied.Methods Data mining analysis of the association between monitoring data for multiple mycotoxins contamination in wheat was performed using the association rule Apriori algorithm. Boolean data type of transaction database was constructed according to the pollutant index values to risk hierarchy structure to mine frequent item sets of transaction database. To determine frequent item sets and obtain strong association rules， minimum threshold support and minimum confidence was set， and iterative connection and pruning operations were performed repeatedly. Association rules were evaluated by confidence， support and promotion degree， etc. Finally， data visualization was applied to association rules to display and verify rules more intuitively.Results The potential strong association rules of co-contamination of muti-mycotoxins in wheat were found， including 9 strong association rules of single common contamination toxin and several strong association rules of combined term sets. The co-pollution relationship between deoxynivalenol and nivalenol， zearalenone and deoxynivalenol was analyzed and verified. The confidence was 92.0% and 80.6%， respectively.Conclusion The strong association rules obtained by data mining have certain significance for the early warning， prevention and control of wheat toxin risk， which provides basis for the assessment of combined exposure to multiple toxins.