Young firms have long been known to be important drivers of job creation, yet little is known about how and who these firms hire. We provide new evidence that they are disproportionate employers of a largely “unemployable” group: individuals with criminal records. Using administrative data on U.S. workers, firms, and the criminal justice system, we show that young firms struggle to compete for labor against established firms and subsequently hire disproportionately from people with criminal records. Following a criminal conviction, a worker becomes 21% more likely to take a job at a young firm. We show this pattern reflects firm-side constraints rather than worker or young firm preferences: young firms conduct fewer background checks and draw from a residual pool of workers that established firms have screened out. Consistent with this mechanism, sorting is substantially stronger for heavily stigmatized offenses, weaker in tight labor markets where established firms lower their screening standards, and weaker under Ban-the-Box policies that restrict established firms’ ability to screen. Using a Bartik instrument, we show that counties with more jobs at young firms see higher employment rates for workers with criminal records but no change in recidivism, suggesting these jobs expand labor market access for these workers, but may not produce the stability needed to deter re-offending.