Safety is an absolute prerequisite for introducing any new therapy, and the need to monitor the consequences of administration of both vector and transgene to the fetus is particularly important. The unique features of fetal development that make it an attractive target for gene therapy, such as its immature immune system and rapidly dividing populations of stem cells, also mean that small perturbations in pregnancy can have significant short- and long-term consequences. Certain features of the viral vectors used, the product of the delivered gene, and sometimes the invasive techniques necessary to deliver the construct to the fetus in utero have the potential to do harm. An important goal of prenatal gene therapy research is to develop clinically relevant techniques that could be applied to cure or ameliorate human disease in utero on large animal models such as sheep or nonhuman primates. Equally important is the use of these models to monitor for potential adverse effects of such interventions. These large animal models provide good representation of individual patient-based investigations. However, analyses that require defined genetic backgrounds, high throughput, defined variability and statistical analyses, e.g. for initial studies on teratogenic and oncogenic effects, are best performed on larger groups of small animals, in particular mice. This chapter gives an overview of the potential adverse effects in relation to prenatal gene therapy and describes the techniques that can be used experimentally in a large animal model to monitor the potential adverse consequences of prenatal gene therapy, with relevance to clinical application. The sheep model is particularly useful to allow serial monitoring of fetal growth and well-being after delivery of prenatal gene therapy. It is also amenable to serially sampling using minimally invasive and clinically relevant techniques such as ultrasound-guided blood sampling. For more invasive long-term monitoring, we describe telemetric techniques to measure the haemodynamics of the mother or fetus, for example, that interferes minimally with normal animal behaviour. Implanted catheters can also be used for serial fetal blood sampling during gestation. Finally, we describe methods to monitor events around birth and long-term neonatal follow-up that are important when considering human translation of this therapy.