Although frowned upon, bridal pregnancies were a common phenomenon in early twentieth-century Netherlands, as they were in many European countries. However, there was a marked regional and social variation in their occurrence. This variation still awaits explanation. In this period with its ineffective forms of contraception, sexuality before marriage was risky. Who took those risks and in what circumstances? What cues from the environment 'allowed' (or refrained) unmarried couples from having sex? And what is the relevant 'environment' for adolescent lovers: what was the role of their peer group, of their parents, and of their church? In this article, we have used a dataset of nearly ten thousand fertile marriages (contracted between 1870 and 1950) with detailed information on the family backgrounds of the spouses. Moreover, we have combined this dataset with a questionnaire on local courtship customs in the early twentieth century. This allows us to perform a multilevel regression analysis of the likelihood of a premarital pregnancy, in which we can look simultaneously at the effects of social class, religion, family composition, characteristics of the couples themselves (level of homogamy), and local courtship customs. Our analysis of courtship customs focuses on communal norms about the timing of courtship (how young could one start), but especially on local practices regarding who (if at all) supervised the meetings of the lovers. Our results confirm earlier findings that bridal pregnancy in The Netherlands was strongly concentrated in proletarian as well as in protestant groups. We find evidence for parental tolerance for sexual urges of (endogamous) young couples who posed no threat to the planned property transmission. However, we also find evidence that youths deliberately advanced a marriage (using a pregnancy as leverage) to gain independence. Local and regional courtship customs seemed not strongly associated with high or low levels of bridal pregnancies.