This paper describes a Bayesian approach to source localization and tracking with uncertain ocean environmental parameters (water column and seabed). The inversion is formulated for both source location and environmental parameters, and solved using Gibbs sampling methods which sample directly from the posterior probability density. The information content for source localization is quantified in terms of probability ambiguity surfaces, consisting of joint marginal probability distributions for source range and depth integrated over unknown environmental parameters, similar to the optimum uncertain field processor (OUFP). The posterior environmental information content, including both prior information and information provided by the acoustic data, is also quantified in terms of marginal distributions. In source tracking, knowledge of the environmental parameters can be augmented as the tracking progresses, by explicitly passing posterior environmental information from one localization on, as improved prior information in the subsequent localization. Information on interparameter correlations is included by formulating marginal distributions in terms of principle components (empirical orthogonal functions) of the environmental parameters.