The Box-Cox transformation is used to make data more suitable for statistical analysis. Its application to extreme value statistics increases the convergence rate of some classical estimators of the tail index and reduces their bias in the context of independent and identically distributed (i.i.d.) random variables. In this paper, we study a bias-reducing estimator of the extreme value index under β-mixing serials dependence using the Box-Cox transformation. Under certain regular conditions, we establish the asymptotic normality of the proposed estimator and derive an asymptotically unbiased estimator of high quantiles. In a simulation study, we highlight the performance of our proposals, which we compare with alternative estimators recently introduced in the literature.