An alternative empirical method to estimating the labour supply function is proposed, based upon subjective wellbeing data. It potentially addresses limitations of the standard neo-classical approach by allowing workers’ observed hours worked to deviate from their utility maximising point.