This paper examines the earnings returns to learning that takes place following the conventional "school-to-work" stage of the life-course. We operationalise such "lifelong learning" as the attainment of certified qualifications in adulthood, following the completion of the first period of continuous full-time education. Using data from the British Household Panel Survey (BHPS) for the period 1991-2006, our approach and findings represent an important addition to the existing evidence base. By using annual data, we are able to employ the fixed effects estimator, which eliminates the problem of time-invariant unobserved heterogeneity. Our dynamic specification uses a lag structure to consider how earnings returns evolve in the medium and longer run, whilst also controlling for wage trends which were evident "prior" to qualification attainment. Our results show a medium-run return for women of 10% on hourly wages. For men, initial suggestions of a similar positive return are eliminated once pre-qualification trends are taken into account. This suggests that adult learning has a causal effect on women's subsequent earnings but, for men, any apparent gain is due to selection. (Contains 6 tables and 3 figures.)