In this paper, we investigate the parametric component and nonparametric component estimators in a semiparametric regression model based on m-asymptotic negatively associated (m-ANA, for short) random variables. The r-th (r > 1) mean consistency, complete consistency and uniform consistency are obtained under some suitable conditions. In order to assess the finite sample performance, we also present a numerical simulation in the last section of the paper. The results obtained in the paper extend the corresponding ones for independent random errors, φ-mixing and other dependent random errors.