Plastic tearing occurs in compact tension (C(T)) specimens fabricated from ductile steels. This tearing, however, is not an elastic fracture problem governed by a crack tip singularity parameter. The current application of the J-integral and the J-R curve to plastic fracture mechanics is misleading. This dissertation reviews the existing energy rate and geometric factor approaches, and then reanalyzes the unloading compliance data for C(T) specimens of various types of steel. The objective of the analysis is to investigate the criteria that can characterize the crack extension under plastic tearing. The author suggests that the energy release rate that remains constant with significant crack growth, in conjunction with the mechanism motion of crack extension to represent the degree of tearing, can serve as the objective parameter in a standard fracture toughness test. Both the energy release rate and the plastic tearing mechanism motion are calculated directly from the test data. The effects of the initial crack length, the specimen size, the testing temperature, and the steel type on the results are investigated. The purpose is to determine the sensitivity of both parameters to the testing configurations. To discuss the tri-axial stress status ahead of the crack front, a finite element model using non-singular elements is presented and verified by the experimental load-displacement curve. Simplifications to the costly C(T) specimen fabrication are proposed based on the analytical conclusions from both the test data and the finite element model.