Current practice in antenatal care aims to assess possible risk factors which may affect the mother, the fetus or both. These risks are evaluated in general (as for maternal age or parental ethnic background), and also specifically (for risks identified through family histories and the history of previous pregnancies). Once these data are ascertained and discussed between parents and doctors, choices involving further investigation of the fetus may need to be made. This may involve invasive procedures, such as amniocentesis or chorionic villus biopsy which may pose a risk (albeit usually small) to the pregnancy. It has been known for many years that fetal cells escape into the maternal blood. However, it is only in the last few years that new molecular techniques have been developed which could utilise these cells for non invasive prenatal diagnosis. This thesis investigates the possibility of gaining information about the pregnancy from fetal cells shed into the maternal circulation. I first conducted a prospective study to determine the time of appearance of fetal cells in the maternal circulation. Using PCR analysis of maternal blood samples, fetal DNA was detected in all the pregnancies studied at seven weeks and the fetal sex correctly predicted. These data demonstrate that fetal DNA is present in all pregnancies at a time suitable for first trimester prenatal diagnosis, that it first appears within a two week window, and that (as DNA is no longer detected two months after delivery), contamination with DNA from previous pregnancies should not be a problem. I next investigated the use of PCR to assess quantitatively the amount of fetal DNA directly on whole blood samples, since the routes of fetal cell dissemination suggest that the amount may increase during pregnancy and may be linked with the development of maternal raised blood pressure and intrauterine growth retardation. The PCR method chosen was found to be unsuitable for quantitation of the small amounts of fetal DNA present in the vast excess of maternal DNA. Enrichment of cells from the maternal circulation will be necessary before DNA analysis or karyotyping can be carried out easily. Antibodies to the epidermal growth factor receptor were attached to magnetic beads which were then used in test systems of first trimester placenta and blood to attempt to separate trophoblast with successful enrichment of 150-fold. The system was then applied to maternal blood samples and the isolated cells analysed with dual fluorescence in situ hybridisation, to detect male (fetal) cells. A system for the isolation of a single interphase nucleus was then developed, to permit single cell analysis of single gene disorders. The nuclear X and Y chromosomes were identified with dual fluorescence in situ hybridisation and the nucleus was removed from the slide with a micro-dissection needle. The DNA was co-amplified using two sets of PCR primers for the detection of amelogenin and a common mutation causing cystic fibrosis. Encouraging results were obtained, and further refinement of the technique for single nuclear isolation could provide non-invasive diagnosis for a single gene disorder. However, close attention to the avoidance of DNA contamination will be required to provide reliable results.