The kernel estimators of nonparametric regression function are considered in this article Under weaker conditions, the law of interated logarithm for the kernel estimation of nonparametric regression function estimators are discussed by the means of the error decomposition of kernel estimator