A large literature in behavioral and labor economics documents gender differences in personality traits and preferences, as well as their explanatory power for gender gaps in occupational choice and career success. These studies usually focus on a single trait or personality classification, such as competitiveness, risk preferences, or the Big Five personality inventory. In this paper, I instead ask how much of gender differences in occupational sorting can be statistically explained by a comprehensive range of trait and preference measures jointly. I combine detailed personality and preference indicators elicited in a representative Dutch survey panel and link them to career outcomes for which large gender gaps are observed: the underrepresentation of women in management and math-intensive occupations, and the underrepresentation of men in teaching and caring occupations and the public sector. Correcting for measurement error, differences in preferences and personality can statistically explain a large part – typically half or more – of gender differences in occupational sorting. Traits with a "dark" side – such as willingness to play dirty, externalizing behavior or psychopathy – capture a surprisingly large share of these gaps.