Many studies show that motherhood has substantial impacts on women's wages and earnings, but there is less evidence on the effect of the timing of entry into motherhood, particularly over the long term and from contexts other than the US. We analyse a sample of women who became mothers by age 30 from the 1970 British Cohort Study to examine whether the timing of the first birth affects mothers' midlife earnings, as well as the role of potential mediators. In the framework of instrumental variable regression, our preferred specification utilizes the occurrence of contraceptive failure as a source of exogenous variation in the age at first birth. We find tentative evidence that a later first birth leads to a higher probability of having any earnings in midlife. However, a later first birth has a negative earnings effect among mothers with any midlife earnings. Potential mediators of this effect are part-time work and birth spacing, given that a later first birth led to a higher likelihood to work part-time at midlife and to a shorter interval between the first and second birth. These findings suggest that the long-term economic impacts of having an early first birth may not be uniformly negative.