Charpy-V test results are widely used for the surveillance program of RPV (Reactor Pressure Vessel) embrittlement by neutron irradiation. Service life extension of nuclear power plants and more stringent safety requirements are increasing the request for small test specimens such as subsize Charpy. Furthermore, the empirical correlation formulas between conventional Charpy-V and fracture toughness are often questionable. Therefore, prior to use of reduced size specimens, different hypotheses have to be validated and this can only be achieved by combining tests and finite element simulations. The material of interest is a nuclear RPV low alloy ferritic steel 16MND5 (eq. A508 C13). Instrumented impact tests were performed on Charpy-V and subsize Charpy over a wide range of temperatures. The hyperbolic tangent curve from W. Oldfield [1] was fitted to the results. Then, correlation formulas are evaluated and commented. In a first step, the aim of the finite element simulations is to get a clear description of the global mechanical behavior of the specimens. A complete description of dynamic impact tests is not required for fracture mechanics purposes. A quasi-static simulation taking into account strain rate effect on the material response is sufficient. However, a 3D analysis is required even for cleavage failure mode. The second step is related to the transferability of fracture criteria. This task has been initiated at low temperatures using the Beremin cleavage model [2] and in the ductile regime using the Rousselier porous model [3]. The finite element results are compared to the experimental one. These results show good transferability potential.