We analyze how much a rational-choice model can explain the temporal and spatial variation in fertility and school enrollment in France during the 19th century. The originality of our approach is in our reliance on the structural estimation of a system of first-order conditions to identify the deep parameters. Another new dimension is our use of gendered education data, allowing us to have a richer theory that includes mothers, fathers, boys and girls. Results indicate that the rational-choice model explains 38 percent of the variation of fertility over time and across counties, as well as 71 percent and 83 percent of school enrollment of boys and girls, respectively. The analysis of the residuals (unexplained by the economic model) indicates that additional insights might be gained by considering cross-county differences in family structure and cultural barriers.