This paper presents nonparametric statistical estimation of distribution density. The Monte Carlo method is used to show the effects of kernel function for multimodal kernel density estimation. Here it is shown that the novel kernel function is effective for asymmetrical heavy tails distributions.