The current provisions used in the U.S. Code of Federal Regulations for the determination of the fracture toughness of reactor pressure vessel steels employs an assumption that there is a direct correlation between K_Ic lower-bound toughness and the Charpy V-notch transition curve. Such correlations are subject to scatter from both approaches which weakens the reliability of fracture mechanics-based analyses. In this study, precracked Charpy and smaller size specimens are used in three-point static bend testing to develop fracture mechanics based K_Jc values. The testing is performed under carefully controlled conditions such that the values can be used to predict the fracture toughness performance of large specimens. The concept of a universal transition curve (master curve) is applied. Data scatter that is characteristic of commercial grade steels and their weldments is handled by Weibull statistical modeling. The master curve is developed to describe the median K_Jc, fracture toughness for 1T size compact specimens. Size effects are modeled using weakest-link theory and are studied for different specimen geometries. It is shown that precracked Charpy specimens when tested within their confined validity limits follow the weakest-link size-adjustment trend and predict the fracture toughness of larger specimens. Specimens of smaller than Charpy sizes (5 mm thick) exhibit some disparities in results relative to weakest-link size adjustment prediction suggesting that application of such adjustment to very small specimens may have some limitations.