Introduction: Conventional ultrasound has been used in the management of early pregnancy complications. Common clinical situations that are encountered include pregnancies of unknown location (PUL) defined as a woman with a positive pregnancy test, but when no intra or extra-uterine pregnancy is visualised on transvaginal sonography (TVS); intra-uterine pregnancies of uncertain viability (IPUV) defines as when an intrauterine gestational sac (<20mm mean diameter) is seen on ultrasonography with no visible yolk sac or embryo, or an embryo of <7mm crown- rump length with no visible heart activity; and recurrent pregnancy losses (RPL). Emerging evidence suggests that novel strategies may be developed to predict early pregnancy complications such as miscarriage using B mode ultrasound imaging and advanced three-dimensional 3D technology, alongside early pregnancy biochemical parameters. These strategies can be assessed prior to conception, at implantation and subsequently in the early stages of pregnancy. In this thesis we describe a series of studies to evaluate the role of ultrasound and biochemical markers for the prediction of early pregnancy complications in relation to these three time periods. Methods: Pre-conception, we used 3D ultrasound to study longitudinal changes in endometrial morphology and vascularity, the influence of physical injury to the endometrium on these parameters and their relationship with implantation success. We investigated the relationship between the ovulation-implantation (OI) time interval and subsequent implantation or early pregnancy failure. Post-conception, we studied PUL and IPUV. In the PUL group, ultrasonography in combination with serum biochemical markers were used to validate the role of mathematical logistic regression models for predicting pregnancy location and viability. We investigated the performance of biochemical markers in urine to assess whether measurements of urine markers can replace serum markers for the management of PUL. In the IPUV group, we established ultrasound criteria that can be used to definitively diagnose miscarriage. In addition, we studied the growth of these pregnancies and its relationship with pregnancy viability and the appearance of early embryonic structures in the gestational sac. Results: We have provided evidence to support the view that endometrial biopsy enhances endometrial vascularity in women with a history of RPL. We have suggested a safe approach to characterise PUL and have shown that logistic regression-based prediction models can reliably be used and further developed to predict the outcome in PUL, independent of the timing of serum biochemistry. In addition, we have shown a potential role for urine pregnancy biochemistry in the management of PUL. We have developed new reliable definitions to diagnose miscarriage, which have now been adopted both in the UK and internationally. Furthermore, we found that, gestation sac growth is not a safe indicator of pregnancy viability, although cutoff values for embryo growth can be suggested below which a miscarriage can be predicted. Conclusion: The studies described in this thesis have introduced a number of novel findings in relation to the use of ultrasound and biochemistry for predicting, and optimising the management of, early pregnancy complications from the preconception phase, through early implantation, to the post conception phase.