One of the more consistent empirical findings of household survey-based fertility research in developing countries is the positive relationship between landholding and fertility. Despite some differences in interpretation, most analysts regard differentials in fertility by landholding status as reflecting purposive, optimizing reproductive behaviour based on differential demand for children and the psychic, social, and/or economic services they provide to parents. This paper presents a critique of the literature on landholding and fertility, and questions the logic of the standard interpretations. Rather than reflecting differences in demand and deliberate fertility control, the landholding-fertility relationship is, in many settings, more likely an unintentional by-product of behaviour such as temporary labour migration, whose incidence varies by landholding status, and which, through separation of spouses, may produce non-trivial differences in fertility. A case is made for more institutionally sensitive analysis.