In this paper we consider the problem of mobile robot localization with range sensors in outdoor environments. Our approach applies a particle filter to estimate the full six-dimensional state of the robot. To represent the environment we utilize multi-level surface maps which allow the robot to represent vertical structures and multiple levels in the environment. We describe probabilistic motion and sensor models to calculate the proposal distribution and to evaluate the likelihood of observations. Experimental results obtained with a mobile robot in an outdoor environment indicate that our approach can be used to robustly and accurately localize an outdoor vehicle. The experiments also demonstrate that multi-level surface maps lead to a significantly better localization performance than standard elevation maps.