Recent studies posit that mental toughness is a relevant construct for predicting achievement outcomes in academic settings. It is a multidimensional construct that encapsulates psychological resources that facilitate consistent performance despite stressors and challenges. However, recent evidence has called into question its multidimensional aspect. The first purpose of this study was to verify, using a bi-factor model, if mental toughness can be operationalized by (a) multiple dimensions, (b) a general factor, or (c) both a general factor and multiple dimensions. The second goal was to test the nomological validity of the construct. Specifically, we verified whether the specific factors predict, beyond the general factor, academic achievement and preference for difficult tasks. Using a correlational cross-sectional design in which 515 high school students (58.8% girls; M age = 15.68; SD = 1.05) were asked to complete a questionnaire, we found that mental toughness is best conceptualized by a general factor. More specifically, most loadings are higher on the mental toughness general factor than on the specific dimensions. Furthermore, the mental toughness general factor predicts better school achievement and preference for difficult tasks than the specific factors. The results are discussed in terms of their implications for theory and practice.