The probabilistic fracture mechanics(PFM) is a useful analytical tool to assess the integrity of reactor pressure vessel(RPV) at the event of pressurized thermal shock(PTS). In PFM, the probabilities of flaw initiation and propagation are estimated by comparing the applied stress intensity factor with the fracture toughness calculated by the simulation of various stochastic variables. It is known that the results of PFM analyses are dependent on the choice of the stochastic parameters and assumptions. Of the various variables and assumptions, we investigated the effects of the RT$_{NDT}$ shift equations, fracture toughness curves, and flaw distributions on the PFM results for the three PTS transients. The results showed that the combined effects of the RT$_{NDT}$ shift equations and fracture toughness curves are complicated and dependent on the characteristics of the transients, the chemistry of the materials, the fast neutron fluence, and so on.