This paper uses data from the British Household Panel Survey (BHPS) and the British 1970 Cohort Study (BCS70) to investigate the family background and childhood factors that are associated with having a child as a teenager. The advantage of combining results from these two sets of data is that the BHPS analyses are restricted to a few background factors while the BCS70 analyses have far more. However, the results obtained from the BHPS data are reasonably replicated with the BCS70 data in that family social class and having lived with one parent during childhood are significantly associated with a higher likelihood of a teenage birth. From the BCS70 data we show that the effect of having lived with one parent is not significant once child-specific variables, such as self-esteem and teacher rated behaviour, are included in the models. Mothers age at the birth of the cohort member and mothers education have significant, consistent and robust associations with the likelihood of teenage birth. The analyses reported in this paper are part of a larger programme of work for the Department of Health examining the medium and long-term consequences of early childbearing.