Some studies, mainly in the older literature, observed a significant association between miscarriages and birth defects (BDs) occurring in the same sibship. However, few studies examined the BD/miscarriage relationship in depth. In addition nothing has been added to the underlying mechanisms possibly linking both events. The purpose of this work was to identify specific BDs associated with maternal miscarriages. In particular it examined whether the risk depended on the number of losses, and it did suggest the existence of specific factors for each BD/miscarriage association observed. Methods: The study relied on the Latin American Collaborative Study on Congenital Malformations (ECLAMC) database registries including 26,906 live and stillborn infants with one out of 19 selected isolated BDs and 93,853 normal controls. Infants born to primigravid mothers were excluded from the present study. Demographic and reproductive variables were compared between control mothers With and Without previous miscarriages. The number, frequency, and distribution of miscarriages were observed for each BD and controls. A conditional logistic regression was computed to evaluate the miscarriage risk for each BD. Results: Control mothers with previous miscarriages were older, had had more pregnancies, and were less educated. Three risk patterns of miscarriages were observed: a very high risk of miscarriages associated with gastroschisis, omphalocele, and talipes; only one miscarriage associated with spina bifida, and two or more miscarriages associated with hypospadias. Conclusion: These three patterns suggest that different factors underly each BD/miscarriage association: infertility for hypospadias, vascular disruption for gastroschisis and talipes, while for spina bifida, the much debated trophoblastic cell residue theory could not be discarded.