The question of how field of study at university predicts later earnings is of great theoretical and practical relevance. However, when investigating the role of field of study for earnings, researchers often rely on descriptive data with only a few control variables. The present study went beyond such estimates by investigating how field of study relates to earnings when holding important psychological and sociological constructs at the end of high school constant. The analyses were based on a unique longitudinal dataset of 1,162 German high school graduates followed for 14 years, which includes measures of abilities, personality, and family background. Structural equation models revealed significant earnings premia for the fields of engineering/technology, medicine/health, and business/economics even after controlling for abilities, personality, and family background. Including those control variables reduced the raw field of study effects.