This study explored the effect of wage on fertility rate for Ghanaian women with labour supply as a mediating variable. Using data from GLSS 6, it employed the Marginal Mediation Analysis, which combined average marginal effect and appropriate estimation techniques, to decompose the total effect into direct and indirect effects. Endogeneity was inherent in the mediation model and therefore respondents’ belongingness to a trade union was used as an instrument for wage to correct for the anomaly. Two different instrumental variable estimation methods, namely the two stage least squares (2SLS) and the two stage residual inclusion (2SRI), were used after which the average marginal effects of the estimated coefficients were computed. The 2SLS was appropriate for the equation which had labour supply (hours of work) as its dependent variable because the regressand is a continuous variable. In contrast, the equations whose dependent variable was number of children were estimated by the 2SRI due to the nonlinearity of their measured variable. It was revealed that a percentage increase in wage directly reduces the number of children per woman by 0.85. Also, a percentage rise in wage decreases the number of children per woman by 0.016 through labour supply. Thus empowering women in terms of both their earnings and job opportunities on the labour market is effective in combating high fertility rate among Ghanaian women. The more lucrative the job market is, the greater the opportunity cost of home production activities and hence fertility rate will drop.