[en] The multiple-specimen method (low-blow or stop-block tests) is the conventional approach for measuring the upper shelf fracture toughness of metallic materials under impact loading rates, typically fatigue precracked Charpy specimens tested on an instrumented pendulum machine. The method is fairly simple but requires a relatively large number of specimens. Nowadays, several single-specimen methods are available, which are purely based on the analysis of the instrumented force/displacement trace; they don't need any dedicated instrumentation for the measurement of crack extension during the test. Three of these techniques have been applied in this work to low-blow tests performed at different temperatures on two significantly different RPV steels (20MnMoNi55 and JSPS): the Normalization Data Reduction (NDR) technique, Schindler's Analytical 3-Parameter Approach and Chaouadi's method. Analyses have been performed after applying a double fitting approach to the raw test data, which allows selecting a limited set of force/displacement data which are representative of the whole instrumented trace. Results show that all three methods provide acceptable accuracy in terms of both ductile crack initiation and resistance to crack propagation (tearing modulus). However, for this type of analysis we recommend the use of the more widely accepted NDR technique, which is described in detail in the ASTM E1820-01 standard (although the limitations on data smoothness presently enforced in the standard seem incompatible with the oscillations of a typical dynamic PCCv curve). (author)