One important step in fracture mechanical research has been thermal shock experiments (TSE). The TSEs are exceptional because many of them had extremely long crack fronts. As a trend, the TSE initiation fracture toughness results form a lower bound to the small specimen data used for material characterization. Presently, work is under way to produce an ASTM testing standard for fracture toughness testing in the transition region. The standard is based on the VTT approach for statistical treatment of fracture toughness data. Key components in the VTT approach are statistical expressions for describing the data scatter and for predicting a specimens size effect, and an expression for the fracture toughness temperature dependence. In order to examine the validity of the statistical size effect assumption, the TSEs were reevaluated with the VTT approach. For all the TSEs investigated, the application of the statistical size correction has had the effect of bringing the material characterization data in agreement with the actual TSE results. Furthermore, the application of a statistically defined size-corrected lower bound toughness estimate, based on the small specimen material characterization data and the master curve, yield a conservative, but realistic, estimate of the TSE failure behaviour.