Abstract
In this article, based on empirical likelihood method and instrumental variable adjustment technique, we propose an adjusted empirical likelihood based statistical inference procedure for varying coefficient partially non linear models with endogenous covariates. Under some mild conditions, the constructed empirical log-likelihood ratio is shown to be asymptotically chi-squared, and then the confidence intervals for the parameter and non parametric components are constructed. Some simulation studies are conducted to examine the finite sample performance of the proposed methods.