Studies demonstrating a relationship between small size at birth and adult cardiovascular disease suggest an improvement in fetal growth may lead to reductions in adult disease. The size and body proportions of the baby at birth are partly determined by maternal body composition. Most studies have only considered maternal height and weight, and their relationships with neonatal birthweight, so a clearer understanding of this area is required. Paternal size and body composition also play a role, primarily through the fetal genome, although few studies have investigated father to baby relationships. This thesis uses a number of datasets to characterise geographical variation in neonatal and maternal phenotypes, and investigate both maternal-neonatal and paternal-neonatal relationships. These include cohorts from UK, Finland, India, Sri Lanka, China, Congo, Nigeria and Jamaica. Analyses were restricted to singleton, livebom, term births. Neonates in Europe were the largest, followed by Jamaica, China then Africa, India and Sri Lanka. There was wide variation in many of the measurements such as birthweight, where the mean values ranged from 2730g to 3570g across populations. However, head circumference was similar in all populations except China, where it was markedly smaller. The main differences between populations were in the ratio of head to length, with small heads in China and large heads in India, Sri Lanka and Africa, relative to length. The mothers from Sri Lanka were the shortest (mean height 151cm) and thinnest (mean BMI at 30 weeks gestation 20 kg/m^), while those from Southampton were the tallest (mean height 164cm) and fattest (mean BMI 27 kg/m^). There were large differences between mothers in the amount of fat relative to muscle. Urban Indian mothers were relatively fat while mothers from the Congo, rural India and particularly Jamaica were relatively muscular. Mother to baby relationships were surprisingly similar across populations, although some effects were stronger in developing countries. All the maternal variables had important effects on the neonatal measures, particularly maternal birthweight. 'Like with like' relationships were seen consistently for maternal height and neonatal length, maternal and neonatal head, and maternal and neonatal fat. Maternal muscle effects were relatively weak, except in one dataset (Congo). After adjusting for the variation in maternal phenotypes across populations, differences in neonatal phenotypes were reduced but still present. Paternal height had the strongest effect on neonatal length, while effects of paternal BMI were generally similar across the neonatal measures. When compared with maternal height and BMI, paternal effects were weaker in most datasets. As maternal body composition was shown to explain a large part of the geographical variation between neonates, and all the maternal variables had independent effects on neonatal phenotype, this implies that nutrition during the whole of the mother's life cycle influences fetal growth, not just her body composition during pregnancy. As paternal size also influenced neonatal phenotype, although to a lesser extent, this is likely to reflect genetic effects, which appear to be stronger for the skeleton than the soft tissues.