Abstract: Existing thixotropy test methods (loading methods) include constant shear rate, shear rate stepping, control hysteresis loops of shear rates, shear stress stepping, and control hysteresis loops of shear stresses, etc.. These methods are used to test the thixotropy of waxy crude oil under the same conditions, and Houska thixotropy model and double-structural parameter thixotropy models are used for fitting. Thixotropy parameters from fitting of test data by one method are used to predict the stress of other test methods under loading conditions. Based on the coincidence degree between prediction results and measured values, the test method is evaluated. Results show that no matter which test method is used, the thixotropy parameters from fitting of test data obtained by the use of loading conditions with high shear rates can all successfully predict stress under loading conditions with lower shear rates, otherwise the structural parameter is negative so as to cause prediction failure; the thixotropy parameters from common fitting of multiple group of test data by the same testing method can improve the coincidence degree between predicted values and measured values of stress under other loading methods.