The empirical connection between financial leverage and equity risk premia is surprisingly weak. We link limited financial flexibility to levered risk premia with a quantitative model of firm dynamics, where firms make investment, financing, and default decisions facing idiosyncratic and aggregate risk. Our model spotlights two key variables, leverage gaps and leverage targets, as drivers of risk premia. Notably, leverage alone offers limited insights. Firms only partially close the gap toward their target, being optimally over- or under-levered. Equity holders of over-levered firms bear relatively higher costs of debt, as their capital structure is vulnerable to bankruptcy costs. Hence, leverage gaps contribute to the amplification of asset return fluctuations. The "lost" leverage risk premium reappears after controlling for targets, which instead lower risk premia. Based on the model predictions, we document novel empirical facts, underscoring limited financial flexibility as a plausible explanation for the contentious evidence concerning levered risk premia.