Finite element and analog techniques are used to assess the influence of abnormal crack morphologies on the potential drop (PD) response of an M(T) specimen. All crack geometries considered, including simply deflected, bifurcated, and periodic sawtooth morphologies, result in higher potentials than observed for an undeflected crack. The greatest influence was observed with a bifurcated crack even with relatively short branch segment lengths. The potential drop responses are assessed in terms of the overestimates of stress intensity factor which result if a standard, undeflected crack PD calibration is applied to the PD response of the deflected geometry. The overestimate of stress intensity factor generally increases as deflection angle or crack length increases. Furthermore the PD response measured from a single probe set configuration is slightly less influenced by crack deflection than a dual probe configuration. This influence is also minimized by increasing the probe gage length and sacrificing overall sensitivity. Provided the deflection geometry of a given crack is constant, an effective probe gage length can be determined and used with the standard undeflected crack calibration to approximate closely the PD response of the deflected geometry.