We provide a comprehensive assessment of the margins along which unions impact workers’ careers. To perform our analysis, we combine exogenous variation in union membership take-up with detailed administrative data and two novel field surveys. We find that the career effect of union membership differs greatly depending on the age at which workers enroll. In addition, we show that focusing on a restricted set of outcomes, such as wages and employment, generates a fractionalized understanding of the multidimensional career effect that union membership has on workers. Finally, we show that individual-level union membership matters above and beyond firm-level union density and that union-provided goods contain a substantial private goods component.