Association often occurs in certain reliability theory problems, as well as in some important models employed in statistical mechanics. In this paper, we investigate the construction of the confidence regions for the regression vector $\beta$ in a linear model under stationary and associated errors. It is shown that the blockwise empirical likelihood (EL) ratio statistic for $\beta$ asymptotically has $\chi^2$ distribution. The result is applied to obtain an EL-based confidence region for $\beta$.