Design of Experiments (DoE) is widely used in design, manufacturing and quality management. The resulting data is usually analysed with multiple linear regression to generate polynomial equations that describe the relationship between process inputs and outputs. These equations enable us to understand how input values affect the predicted value of one or more outputs and find good set points for the inputs. However, to develop robust manufacturing processes, we also need to understand how variation in these inputs appears as variation in the output. This understanding allows us to define set points and control tolerances for the inputs that will keep the outputs within their required specification windows. Tolerance analysis provides a powerful way of finding input settings and ranges that minimise output variation to produce a process that is robust. In many practical applications, tolerance analysis exploits Monte Carlo simulation of the polynomial model generated from DoE’s. This paper briefly describes tolerance analysis and then shows how Monte Carlo simulation experiments using space-filling designs can be used to find the input settings that result in a robust process. Using this approach, engineers can quickly and easily identify the key inputs responsible for transferring undesired variation to their process outputs and identify the set points and ranges that make their process as robust as possible. If the process is not sufficiently robust, they can rationally investigate different strategies to improve it. A case study approach is used to aid explanation and understanding.