In recent years, the development of statistical analysis methods designed for the asymmetric dependence modeling has made exciting and rapid progress.However, one issue for the most prominent statistical measures of association in literature, like Pearsons correlation coefficients, Spearmans ρ and Kendalls τ, is that they are not suitable if the data shows the non-linear relationship between the variabls. Motivated by this, we review a new subcopula-based measure of the asymmetric association in a two-way contingency table,which was proposed by Wei and Kim(2017).We examine the use of the measure in detecting the nonlinear association for data generated from continuous random variables. The procedure is developed as a non-parametric approach and assumes no parametric forms for the associated random variables.The proposed procedure is illustrated through a simulation example.