This paper proposes an explanation for several decades of rising U.S. nonmarital birth rates and shares, and for cross-sectional differences in black-white fertility. Significantly, the explanation does not rely on changes over time or differences across races in individual fertility behavior. It is consistent with the rising nonmarital fertility measures observed in the United States since the mid-1970s, higher measured fertility for unmarried blacks than whites, and differences across races in the timing of childbearing, despite nearly constant total fertility rates and increasingly similar target family sizes for blacks and whites. The explanation relies on a selection effect associated with changes in the marriage rate and on racial differences in access to human capital investment opportunities. We find strong support for the explanation using U.S. data over the period 1957–2002. Our findings suggest caution in interpreting the results of empirical studies of childbearing that examine marital and nonmarital fertility rates separately, as these studies typically ignore the selection effect of marriage. (JEL J12, J13, I38)