Artemisinin derivatives are clinically effective and safe antimalarials, but are not recommended during the first trimester of pregnancy because of the resorptions and abnormalities seen in animal reproduction studies. Understanding how, when and what toxicity occurs is crucial to any assessment of clinical relevance. Previously, DHA has been shown in the rat whole embryo culture (WEC) to primarily affect primitive red blood cells (RBCs) causing subsequent tissue damage and dysmorphogenesis. To verify the primary target of DHA in vivo and to detect consequences induced by early damage on embryo development, pregnant female rats were orally treated on gestation days (GD) 9.5 and 10.5 with 7.5 or 15 mg/kg/day DHA and caesarean sectioned on GD11.5, 12.5, 13.5, 15 and 20. A parallel in vitro WEC study evaluated the role of oxidative damage and examined blood islands and primitive RBCs. In accordance with the WEC results, primitive RBCs from yolk sac hematopoiesis were the target of DHA in vivo. The resulting anemia led to cell damage, which depending on its degree, was either diffuse or focal. Embryonic response to acute anemia varied from complete recovery to malformation and death, depending on the extent of cell death. Malformations occurred only in litters with embryonic deaths. DHA induced low glutathione levels in RBCs, indicating that oxidative stress may be involved in artemisinin toxicity; effects were extremely rapid, with altered RBCs seen as early as GD10. In establishing the relevance of these findings to humans, one should consider differences in the development of rodents and humans. While yolk sac hematopoiesis occurs similarly in the two species, early placentation and extent of exposure differ. In particular, early hematopoiesis takes only 7 days in rats (during which RBCs expand in a clonal fashion) compared with 6 weeks in humans; thus the susceptible period in relation to the duration of exposure to an artemisinin-based treatment may be substantially different.