Let
$\boldsymbol{Y}=\left(Y_1, Y_2,\dots, Y_k\right)'$ be a random
vector with multinomial distribution. In this paper we investigate
the convergence rate of so-called power divergence family of
statistics $\{I^\lambda(\boldsymbol{Y}),\lambda\in\mathbb{R}\}$
introduced by Cressie and Read (1984) to chi-square distribution.
It is proved that for every $k\ge4$
$\Pr(2nI^\lambda(\boldsymbol{Y})