ABSTRACTThe ductile–brittle transition temperature is not intrinsic to steel but depends on the specimen type and loading mode used for the test. The influence of these parameters is related to the plastic constraint. Here, the constraint is evaluated by the effective T‐stress obtained by the stress difference method on the notch tip stress distribution. From Charpy energy values at different temperatures, it is possible to obtain a material failure master curve where the notch fracture toughness is plotted versus the shift of test temperature with the transition temperature corresponding to the effective T‐stress value. Values of the critical effective T‐stress are obtained for three subsize Charpy specimens (3/4, 1/2 and 1/3) and are reported on a material transition temperature master curve versus effective T‐stress. This allows obtaining the shift of transition temperature when comparing with those of a standard Charpy V specimen. The results do not confirm the prediction given by the empirical relation of the ASME.