The continuous increment of the required design lifetime for many machinery components and the experimental evidence that materials could fail at stress amplitudes below the conventional fatigue limit have led to a growing interest in the study of the Very-High-Cycle Fatigue (VHCF) behavior of materials. Experimental results showed that materials in the VHCF regime could fail generally due to cracks nucleating around defects within the material (internal nucleation). Therefore, it is generally acknowledged in the literature that the VHCF response of materials is strongly affected by the defect population and, in particular, by the characteristic defect size, which statistically increases with the material volume. According to this well-known dependency, experimental results showed that the larger the loaded volume (risk-volume), the smaller the VHCF strength (size effect). However, a significant increment of the risk-volume is not possible with common specimen shapes used for ultrasonic tests. As a consequence, size effect at large risk-volumes is generally statistically predicted, without any experimental validation. The thesis proposes a new experimental approach for the assessment of size effect in the VHCF regime. Fully reversed tension–compression tests were carried out on specimens with significantly different risk-volumes (hourgalss and Gaussian specimens) by using the ultrasonic testing machines developed at Politecnico di Torino during the Ph.D. The experimental results were analyzed by considering and integrating the well-known statistical models proposed in the literature. A methodology for obtaining reliable and conservative predictions of the VHCF response of materials at different risk-volumes is proposed and discussed in the thesis. Finally, a simple numerical-experimental procedure for the design of components subjected to VHCF failures and characterized by large risk-volumes is proposed.