As path loss modeling for wireless channel is an essential issue for applications of wireless sensor network in wheat fields, 4 kinds of path loss modeling methods were studied and their adaptability was analyzed. Firstly, path loss of 2.4GHz wireless signals was measured by wireless sensor modules in wheat fields, and the trend of path loss was analyzed. Then the measured data were compared with the path loss predicted by the free space model and the two ray model respectively. A modified two slope log distance model was proposed and then the one slope log distance model and the modified two slope log distance model were applied to perform regressions for the measured data of path loss based on the least square method respectively. The results show that there were big differences between the path loss predicted by the first two models and the measured data, and thus the first two models couldnot be used to predict path loss directly in the environment. The R2 values of the latter two models were 0.879 and 0.943 and the standard error of the estimate were 3.242 and 2.229 respectively, which indicated both of the two models were applicable for path loss modeling in wheat fields and the modified two slope log distance model had a better accuracy.