Abstract: We show that bank credit affects entrepreneurship, but only in low-income regions. We use a novel methodology to identify credit supply shock from regional demand shock using comprehensive data on small business loans between pairs of banks and counties in the US. While there is no impact in top income quartile counties, we document that a one std credit supply shock is associated with 1.6 and 1.7 percentage point employment and payroll growth in newborn firms in bottom income quartile counties. We show that this impact is long-lasting; is pronounced only in newborn firms; is not just a redistribution of labor from established to newborn firms; and does not follow with a reduction in labor productivity. We estimate that a credit redistribution of $100 from counties of above- to below-median income per capita results in $6.5 annual labor earnings in the aggregate.