Human pregnancy disorders such as miscarriage, pre-eclampsia, intrauterine growth restriction and adherent placenta are thought to have their origins in defective trophoblastic invasion of the decidua. However, an elucidation of the exact mechanism of invasion is elusive because of poor understanding of early trophoblastdecidua interactions. Research is hampered by the inaccessibility of site of implantation, ethical constraints on studies of human pregnancy and lack of an adequate animal model. This research has sought to overcome many of these challenges by modelling, in vitro, events at the trophoblast-decidua interface. Human trophoblast is relatively readily available from pregnancy terminations or term pregnancies and cell lines have been produced in our laboratory. Decidua has previously proven more challenging to obtain in pure form. During this research a novel direct-vision hysteroscopic technique of decidual biopsy was developed which allowed accurate directed sampling of decidua parietalis separate from decidua basalis. Morphological and immunohistochemical studies confirmed the accuracy of the samples. Further characterisation of pure decidua parietalis versus basalis identified structural and functional changes likely to have been induced by or in response to trophoblastic invasion. For example, using electron microscopy, thicker and disrupted collagen fibrils were noted in decidua basalis versus uniform arrangement in decidua parietalis. Proteomic analysis established basic protein profiles for each type of decidua, and among the differences noted were 34 proteins with differential expression between basalis and parietalis. Consistent with the ultrastructural differences, extracellular matrix component lumican showed differences in expression between the two types of decidua. In vitro explant co- culture models were established with placental villi and pure decidua parietalis biopsies. These demonstrated early evidence of invasion of decidua parietalis by extravillous cytotrophoblast, with viability up to 96 hours. It is now possible, using the tools developed during this research, to conduct in-depth studies of early events at the materno-fetal interface.