Background: In many European countries, the rate of childlessness has increased and this increase is disproportionately large among men. It is thus important to study the antecedents of childlessness. Intelligence as a predictor of childlessness has received some attention, although the findings are conflicted. Men with less education, lower wages and no partner seem to be overrepresented among the childless. These factors are known to be associated with intelligence. However, little is known about how these factors may mediate the association between intelligence and childlessness. Method: The present study analyzed data from the Young in Norway survey, which had been linked to the Historical Event Database (FD-Trygd) of Statistics Norway, in addition to intelligence scores from the evaluation for mandatory military service for all men in Norway. Intelligence scores were used as predictors of childlessness. Information on education, income and relationship history was retrieved from both the Young in Norway survey and the FD-trygd register. Mediation analyses with logistic regression analysis were performed in order to examine the direct and indirect effect of intelligence on childlessness. Results: No linear association between intelligence and childlessness was found. However, follow-up analyses showed that higher risk of childlessness was associated with the highest stanine intelligence score (two standard deviations above the normed average), and that this effect was mediated by whether participants had ever been in a relationship by their late twenties. No significant direct effect of the highest stanine intelligence score was found when the question of whether participants had ever been in a relationship by their late twenties was included. Conclusion: The present study revealed a complex association between intelligence and childlessness. The mediated association between the highest stanine intelligence score and childlessness was interpreted as an expression of delayed family formation. It is still unclear whether this delay in family formation will “catch up” if childlessness is measured at an older age, which raises the need for additional studies on the subject.