Working-class Americans are significantly underrepresented in politics. Some argue that they are filtered out during elections. Others argue that barriers to entry, not the elections themselves, keep them out of office. Are working-class candidates less likely to win elections? This question is difficult to answer because we typically only have the occupational history of candidates who win elections. To overcome this limitation, we collect the full occupation history available on campaign websites for all 8,019 candidates who finished first or second in a congressional primary between 2010 and 2022. Using both difference-in-differences and primary regression discontinuity research designs, we find that when a party nominates a working-class candidate, that candidate is no less-likely to lose their election than non-working-class candidates. This is not because individual candidate characteristics don’t matter in elections, or that working-class candidates aren’t different from other candidates in important ways. Instead, we find that working-class candidates perform as well as other candidates despite having fewer positive valence characteristics, (e.g. ivy league education, a graduate degree or previous political experience), receiving a smaller share of general election donations, and being more ideologically extreme. Taken together, these results provide evidence that general elections do not contribute to the underrepresentation of working-class Americans in government.