This paper uses the semi-parametric method for binary choice models developed by Wang and Zhou (1995) to estimate the effect of health on the labour force participation of married women using data from the Panel Study of Income Dynamics (1989 interviewing year). The semi-parametric method is particularly useful if empirical studies involve many explanatory variables and large sample sizes. The estimation results show that the health condition is a good predictor of participation. The results also demonstrate that the easy-to-compute semi-parametric method is practical for empirical studies. Furthermore, the empirical results suggest that both logit and probit estimators may underestimate the effect of wives' health conditions on their labour force participation.