Purpose -To measure the effects of work-sharing arrangements on participants' subsequent labor market outcomes in Taiwan such as full-time employment rates, working hours of women and men and the difference in scale effect and effect of substitution between hours and employment for women and men. Design/methodology/approach -Using the data from Manpower Utilization Survey, we applied the differences-in-differences estimation method to test the effects of work-sharing arrangements on working hours, wage and employment. Multinomial logit was used to measure the effects of work sharing on full-time employment. In order to correct the simultaneity and selectivity problems, we followed the Heckman two-stage selection procedures to solve the selection bias, and used weighted least squares to solve heteroskedasticity in the wage and hour equations. The instrumental variable (IV) method was used to avoid simultaneity bias in the hour equation.Findings -This paper found the restrictions enforced by law on working hours have negative effects on employees' working conditions in certain industries in Taiwan. After controlling the working hours, we found the wages paid to women and men have increased subsequent to the enforcement of law. However, compared to men the net wage earned by women has increased to a lesser extent. It was further observed that with enactment of work-sharing law, the employment rate of women has considerably declined since 2001. Main findings assimilating the results for hour, wage and full-time employment suggest that a country like Taiwan (with work-sharing arrangements implemented by law) has witnessed a smaller gap between women's and men's working time and wages during