A strain gauge technique recently developed with the wedge test, for estimating crack length and, thus, the fracture energy of structural adhesive bonding, has been employed on a system in which one adherend had two types of surface treatment. Simple polishing and polishing with subsequent sandblasting were the treatments used, with a distinct straight line, perpendicular to the sample edges, separating the two. Despite the clear-cut difference in surface treatment, smooth transitions in crack growth speed were noted. This can be explained by the existence of a curved crack front, encroaching gradually on one surface-treated zone, whilst remaining partially on the other. Crack length, a, vs. time, t, curves were exploited to obtain fracture energy vs. crack speed. The multi-valued nature of the relation can also be explained by a non-rectilinear fracture front. The method is proposed as a method for reproducible comparison of surface treatments.