The Chapter will describe a range of mathematical tools for model sensitivity and uncertainty analysis which may assist in the evaluation of large combustion mechanisms. The aim of such methods is to determine key model input parameters that drive the uncertainty in predicted model outputs. Approaches based on linear sensitivity, linear uncertainty and global uncertainty analysis will be described as well as examples of their application to chemical kinetic modelling in combustion. Improving the robustness of model predictions depends on reducing the extent of uncertainty within the input parameters. This can be achieved via a variety of methods including measurements and theoretical calculations. Optimization techniques which bring together wide sources of data can assist in further constraining the input parameters of a model and therefore reducing the overall model uncertainty. Such methods and their recent application to several combustion mechanisms will be described here.