Zika virus (ZIKV) infection during human pregnancy may cause diverse and serious congenital defects in the developing fetus. Previous efforts to generate animal models of human ZIKV infection and clinical symptoms often involved manipulating mice to impair their Type I interferon (IFN) signaling, thereby allowing enhanced infection and vertical transmission of virus to the embryo. Here, we show that even pregnant mice competent to generate Type I IFN responses that can limit ZIKV infection nonetheless develop profound placental pathology and high frequency of fetal demise. We consistently found that maternal ZIKV exposure led to placental pathology and that ZIKV RNA levels measured in maternal, placental or embryonic tissues were not predictive of the pathological effects seen in the embryos. Placental pathology included trophoblast hyperplasia in the labyrinth, trophoblast giant cell necrosis in the junctional zone, and loss of embryonic vessels. Our findings suggest that, in this context of limited infection, placental pathology rather than embryonic/fetal viral infection may be a stronger contributor to adverse pregnancy outcomes in mice. Our finding demonstrates that in immunocompetent mice, direct viral infection of the embryo is not essential for fetal demise. Our immunologically unmanipulated pregnancy mouse model provides a consistent and easily measurable congenital abnormality readout to assess fetal outcome, and may serve as an additional model to test prophylactic and therapeutic interventions to protect the fetus during pregnancy, and for studying the mechanisms of ZIKV congenital immunopathogenesis.