This paper examines the socio-economic consequences of teenage motherhood for a cohort of British women born in 1970. We employ a number of methods to control for observed and unobserved differences between women who gave birth as a teenager and those who do not. We present results from conventional linear regression models, a propensity score matching estimator, and an instrumental variable estimator that uses miscarriage data to control for unobserved characteristics influencing selection into teenage motherhood. We consider the effects on equivalised family income at age 30, and its constituent parts. We find significant negative effects of teenage motherhood using methods that control only for observed characteristics using linear regression or matching methods. However once unobserved heterogeneity is also taken into account, the evidence for large negative effects becomes much less clear-cut. We look at older and younger teenage mothers separately and find that the negative effects are not necessarily stronger for teenagers falling pregnant before age 18 compared with those falling pregnant between 18 and 20, which could further suggest that some of the negative effects of teenage motherhood are temporary.