Abstract This paper investigates how educational and geographic capital are assorted among households in rural Nepal, and how women's marital age may shape this distribution. Our focus on the timing of marriage adds a new dimension to studies of geographies of youth and marital assortment, while our emphasis on the physical and spatial attributes of households leads us to propose the concept of geographic capital, operationalised here as agrarian landholding and access to rural markets. Using data on 17,284 women from rural lowland Nepal, heat tables showed substantial pairing among uneducated spouses, whereas educated men married women with varying levels of schooling, partly because fewer women were educated. Multivariable logistic regression models showed that the odds of marrying an educated man increased substantially for women with secondary education, and vice versa. Educated women were also more likely to marry into households with geographic capital. However, landowning husbands tended to marry younger wives, perhaps because the natal home was prepared to marry daughters earlier in order to access this geographic capital. The youngest‐marrying women were least likely to marry into households with accessibility to markets. Our findings may help understand the decisions of both a woman's natal and marital household over the timing of her marriage, and the investment in her formal education. These patterns have implications for both spouses because capital not only shapes marital pairing, but also the spatial niche of the household within which women and their children will experience their life‐course.