Surface finish of engineering components is measured and controlled to achieve a desired function. Surface finish can also be used to provide feedback to the manufacturing process. An area of active research in surface metrology is in developing tools and techniques for providing greater insight into the relationship between surface texture, a component?s function and the manufacturing process. In this context, this paper explores the use of different clustering techniques such as k-means, ISODATA and neural networks to relate surface metrology data to a component?s function and the manufacturing process that produced the part.