This paper applies semiparametric regression models to shed light on the relationship between body weight and labor market outcomes in Germany. We find conclusive evidence that these relationships are poorly described by linear or quadratic OLS specifications, which have been the main approaches in previous studies. Women's wages and employment probabilities do not follow a linear relationship and are highest at a body weight far below the clinical threshold of obesity. This indicates that looks, rather than health, is the driving force behind the adverse labor market outcomes to which overweight women are subject. Further support is lent to this notion by the fact that wage penalties for overweight and obese women are only observable in white-collar occupations. On the other hand, bigger appears to be better in the case of men, for whom employment prospects increase with weight, albeit with diminishing returns. However, underweight men in blue-collar jobs earn lower wages because they lack the muscular strength required in such occupations.