Because different nodes of analog circuit have different measuring points in response to parameter changes, the selection of measuring points has a great influence on the effect of fault detection. Therefore, a multi-objective selection method of measuring points of analog circuit based on sensitivity factor is proposed. Sensitivity analysis is introduced into the study of the influence of independent variables on dependent variables in analog circuits, and normalized. In the multi-objective selection, the ant colony mechanism is introduced, and the moving direction of ants is guided by the analysis results of sensitive factors, and the sensitive factor matrix is constantly updated to get the optimal measuring point information. In the test results, the FNR value of fault detection corresponding to the design method is the lowest, always stable below 0.06, and the accuracy value is the highest, always stable above 0.95, which is obviously superior to the control group.