The aim of this study was to explore two main hypotheses. Firstly, that the greater similarity in glucose tolerance and blood pressure in monozygotic twins compared with dizygotic twins is due to their more similar prenatal environment. Secondly, that within pair differences in glucose tolerance and blood pressure would be determined by differences in size at birth. A longitudinal study of births in the city of Birmingham from 1950 onwards provided the sampling frame for the study and allowed a population-based sample of twins to be studied. Twins born between 1950 and 1954 were identified and followed up. They were visited at home where their blood pressure was measured. They were then invited to attend a clinic for an oral glucose tolerance test. Adult levels of glucose tolerance and blood pressure were more highly correlated in the monozygotic than dizygotic twins. These trends were not explained by size at birth. Examination of within pair differences in glucose tolerance revealed significant associations between size at birth and insulin resistance in the monozygotic twins. However, inconsistencies in these trends and the fact that they were based on small numbers of twin pairs weakened these findings. Furthermore, the well documented inverse associations between birth size and adult glucose tolerance and blood pressure in singleton populations were absent in the twins. The findings of this study were not consistent with the fetal origins hypothesis. The association of within pair differences in birth size and insulin resistance in monozygotic twins does, however, suggest an effect of birth size that is independent of genetic factors.