To construct and compare the methods for correlation analysis of different elements in wheat, including barium, vanadium, cadmium, lithium, aluminum, manganese, lead, copper, selenium, chromium, and arsenic. Methods Pearson correlation coefficient and partial correlation coefficient were applied and compared to analyze the correlation among the 11 elements in wheat. Results Both of the methods could find the correlations among various contaminants from the data and had their own characteristics on computational complexity, information reflection and other aspects:Pearson correlation coefficient method need less calculation, but also reflected less information; partial correlation coefficient reflected more information but need more samples and computing resources. Under the current data and software and hardware conditions, the correlation analysis of the partial correlation coefficient was proposed.