A previously developed non-conventional J integral based on hysteresis energy and the ASTM E813 methods have been employed to test the fracture toughness of a polycarbonate (PC)/acrylonitrile-butadiene-styrene (ABS) blend. A systematic study of the effect of the specimen thickness (from B=4 to 15 mm) and the presence of a V-shaped side groove on the resultant J(1c) has been carried out. The results show that the J(1c) values obtained from the hysteresis energy method and the E813-81 method are comparable, but are similar to 60-90% lower than those obtained from the E813-87 method. When the specimen thickness is relatively thin, the J(1c) values obtained from the V-shaped-side-grooved specimens are slightly higher than those from samples without the side groove, but this difference gradually decreases with an increase in the specimen thickness. The J(1c) values obtained from the ungrooved specimens are almost independent of the specimen thickness (from B=4 to 15 mm) for all of these methods. The critical initiation displacements in terms of the onset of crack initiation, determined from the plots of hysteresis energy vel sus displacement, hysteresis ratio versus displacement, and crack growth length versus displacement, are fairly close in value. This means that the critical crack initiation and the corresponding J(1c) obtained from this hysteresis energy method indeed represent the actual physical event of the onset of crack initiation.