In Japan, the waitlist for childcare is a serious problem. This problem deters mothers from working, even if they desire to work. One reason for this problem is the regulation of the market for childcare services. This regulation results in childcare prices that are too low, which prevents an increase in supply. Regulations need to be removed to increase the price of childcare and increase its availability. Increasing the price of childcare may burden mothers, but efforts could be made to soften that burden. One method is to increase the wage level for females. In sum, to address the long waitlist for childcare and increase the female labor supply two things are required: 1) increase the price of childcare services, and 2) increase the female wage level. This paper investigates the effects of childcare cost and female hourly wage on Japanese female labor supply and use of childcare service, with data from the 2000s. According to the extant literature, as the price of childcare increases, the female labor supply shrinks and the use of childcare service decreases. However, most of the relevant empirical literature is based on cross-sectional data from the 1990s. It is important to re-examine mothers’ decisions to work and use childcare with data from the 2000s, because regulation of childcare services were changed in Japan, and this may have affected these decisions. Panel data are needed. The data used here were drawn from the “Longitudinal Survey of Adults in the 21st Century (LSA21)” from 2002 to 2012, and the “Survey of Social Welfare Institutions” to incorporate location-specific variables. These data are used to estimate a multinomial logit model that takes into account heterogeneity, sample selection bias, and regional differences. Sample selection bias occurs when data about the price of childcare services and wages of women cannot be observed because there are individuals who do not use childcare services and who do not work. To address this issue, the multinomial logit model is estimated using imputed values derived from a Heckman selection model. The analysis support two conclusions. First, when the price of childcare is high, Japanese mothers choose not to work and not to use childcare services. However, an increase in hourly wages offsets this negative effect. This is especially true for the estimation results using the sub-sample with mothers who worked and used licensed childcare services. Second, in a model using fixed-effects to control for unobserved heterogeneity, the effects of childcare costs on mothers’ decisions to work and to use unlicensed childcare service were insignificant.