The Bayesian inference problem is concerned with estimating parameters in a model given observations. Bayes' Theorem provides a formula for the distribution of the desired parameters. We will discuss how expansions in orthogonal polynomials in random space can allow efficient approximations to this formula.