This paper provides novel evidence on the effect of deregulating overtime hours restrictions on women by using the 1985 Amendments to the Labour Standards Act (LSA) in Japan as a natural experiment. The original LSA of 1947 prohibited women from working overtime exceeding two hours a day; six hours a week; and 150 hours a year. The 1985 Amendments exempted a variety of occupations and industries from such an overtime restriction on women. Applying a difference-in-difference model to census data, we find causal evidence pointing to the positive effect of this particular piece of labour market deregulation on the proportion of female employment. We then carry out a series of sensitivity analyses to ensure the robustness of our finding. Especially, we conduct a falsification test and an event study to show that our causal inference is not threatened by the differential pretreatment trends. Finally, we use quantile regressions and find that for jobs with more rapidly growing proportion of female employment, the effect of the exemption from the overtime work restriction on women is larger.