In previous papers, a protocol using thermal measurements to predict the fatigue lifetime throughout an energy-based criterion afforded a very efficient prediction of the deterministic Wöhler curve with only one sample, within less than 1 day. Nevertheless, these papers investigated a limited number of materials because the fatigue campaigns required to provide a validation of this fast prediction are very demanding in time and specimens. In this study, the PROFEM project (supported by the French National Research Agency) allowed us to investigate a very wide range of materials including natural rubber (NR) and synthetic compounds (SBR). The investigation therefore considers either crystallizing (NR) or noncrystallizing (SBR) materials for several carbon black amounts (0, 20, 43, 58 phr) and types (N220, N326, N375, N550, N772). As a first step, this study presents the application of the fast prediction protocol on one compound and investigates its robustness (repeatability, influence of the N frequency, sensitivity to the graphical analysis). Then, the method is challenged on the 11 other compounds, and the comparison of the predicted fatigue curve with the Wöhler curves obtained by classical campaigns highlights a very good agreement for most of the materials.