Administrative data reveal large spreads in the U.S. unsecured credit market that are in excess of and vary sharply with default risk. Incorporating the empirically correct incidence of these borrowing premia allows standard unsecured credit models to match key granular credit moments, including the distribution of credit balances by default risk. The incidence and response of premia to aggregate shocks influence cross-sectional patterns in and aggregate dynamics of borrowing and default. An extended model infers lending standards from observed shifts in borrowing premia; we find that standards tightened only for risky borrowers at the start of the recession in 2020.