The small punch test (SPT) is very convenient for estimating tensile mechanical properties, being the estimation of fracture toughness still a controversial subject of debate. One of the new strategies developed is the use of notched specimens. In this paper, two different grades of CrMoV steels were employed to analyse the evolution of the notch mouth opening displacement of the small punch sample (δSPT). Complete and interrupted tests were performed on specimens with longitudinal non-through notches with a notch length to thickness ratio of 0.3. A numerical model was also developed for corroborating the experimental results. A material-independent relationship between δSPT and the punch displacement (d) was found: δSPT=0.217d. Since crack length measurement is not possible on SPT samples, the value of δSPT at crack growth onset (δSPT_ini) was used for comparison with the CTOD values for crack initiation in the standard tests (δini). Crack growth onset in the SPT specimens was verified by observation after splitting them in two halves, as well as comparing the numerical curves (without damage model) and the experimental ones. Larger values have been obtained by means of the SPT, due to the lower constraint of the test. However, the developed methodology seems to be suitable when dealing with ductile steels, although other different materials are needed to be tested.