A new approach was implemented to confirm the start of ductile tearing relative to assessments by other methods such as direct-current Electric Potential (d-c EP) method in coupon specimens. This approach was developed on the Key-Curve methodology by Ernst/Joyce and is similar to the ASTM E-1820 Load Normalization procedure used to determine J-R curves directly from load versus Load-Line Displacement (LLD) record of the test specimen. It is consistent with Deformation Plasticity relationships for fully plastic behavior. Using this Experimental Key-Curve method, crack initiation can be determined directly from load versus LLD data or load versus Crack-Mouth Opening Displacement (CMOD) obtained from a fracture test without the need for additional instrumentation required for crack initiation detection. It is based on the fact that plastic deformation of homogeneous metals at the crack tip follows a power-law function until the crack tearing initiates. Crack tearing initiation is determined at the point where the power-law fit to the load versus plastic part of CMOD or LLD curve deviates from the total experimental load versus plastic-CMOD or LLD curve. The procedure for fitting of the data requires some care to be exercised such that the fitted data is beyond the elastic region and early small-scale plastic region of the Load-CMOD or Load-LLD curve but include data before crack initiation. An iterative regression analysis was done to achieve this, which is shown in this paper. The iterative fitting in this region typically results with a coefficient of determination (R2) values that are greater than 0.990. This method can be either used in conjunction with other methods such as direct-current Electric Potential (d-c EP) or unloading-compliance methods as a secondary (or primary) confirmation of crack tearing initiation (and even for crack growth); or can be used alone when other methods cannot be used. Furthermore, when using instrumentation methods for determining crack-initiation such as d-c EP method in a fracture toughness test, it is good to have a secondary confirmation of the initiation point in case of instrumentation malfunction or high signal to noise ratio in the measured d-c EP signals. In addition, the Experimental Key-Curve procedure provides relatively smooth data for the fitting procedure, while unloading-compliance data when used to get small crack growth values frequently has significant variability, which is part of the reason that JIC by ASTM E1820 is determined using an offset with some growth past the very start of ductile tearing. In this work, the Experimental Key-Curve method had been successfully used to determine crack tearing initiation and demonstrate the applicability for different fracture toughness specimen geometries such as SEN(T), and C(T) specimens. In all the cases analyzed, the Experimental Key-Curve method gave consistent results that were in good agreement with other crack tearing initiation measuring method such as d-c EP but seemed to result in less scatter.