The invention relates to the technical field of lightning early-warning, in particular to a method automatically acquiring atmospheric electric field thunder and lightning early-warning characteristic parameter threshold values. The method comprises preprocessing atmospheric electric field monitoring data to acquire all lightning early-warning characteristic parameters; choosing thunder and lightning early-warning characteristic parameters in a history period and actually-happened thunder and lightning data as samples; counting, recording and drawing an association graph between the parameters and thunder and lightning activity, using the association graph to screen out characteristic parameters with strong early-warning distinguishing capacity as basic parameters of thunder and lightning early-warning; setting a group of different threshold values for each parameter according to a range and accuracy of the parameter; evaluating different threshold values of each characteristic parameter according to inspection standards to acquire the best threshold value of each characteristic parameter and finally forming comprehensive characteristic parameter threshold values of atmospheric electric filed thunder and lightning early-warning.