This work studies the stationary distribution of stochastic Lotka-Volterra population dynamics in state-dependent random environments. We first use the stochastic comparison approach to handle the difficulty caused by the state-dependent regime-switching. Then the ergodic property and positive recurrence are investigated by the theory of $M$-matrix under small perturbation.  The mean of the stationary distribution is also estimated. At last, the main results are illustrated by a numerical example. This is a joint work with Huijie Ji.