After considering the tail and outlier of sample, this study estimates Taiwan's leisure expenditure equation based on quantile regression, and examines the sample or sub-sample of the whole family, unclear family and clear family among the population. The empirical result show that the higher level of quantile was one of the most efficiency and more explained in the leisure expenditure equation. By contrast, the linear regressions estimated by the ordinary least square method are unable to reveal the urban, the householder of sex, education, a public official and the ownership of real estate etc. Applying quantile regression to leisure expenditure, this study provides new evidence on Taiwan‘s leisure expenditure, also provides information to researchers, policy makers, financial counselors, and business managers who focus on the leisure expenditures of unclear family and clear family.