Abstract: In the United States, age at first marriage was lowest and the education gap between husbands and wives was highest during the 1950s. The conventional explanation for such a negative correlation between the age at first marriage and the spousal education gap is that early marriage leads to earlier and higher fertility, which in turn prevents women from acquiring education. In this paper, we propose a complementary but novel hypothesis: early marriages enabled couples to overcome some of the credit constraints that could have bound husbands from acquiring education. We show that a calibrated model that includes this motive and mechanism can replicate not only the marriage and education patterns observed in the middle of the century in the United States, but also the trends in both variables over most of the 20th century. We then discuss some further corollary predictions of the theory and show how they match the patterns seen in the relevant data.