The present paper deals with optimisation of Nearest Neighbour rule Classifiers via Genetic Algorithms. The methodology consists on implement a Genetic Algorithm capable of search the input feature space used by the NNR classifier. Results show that is adequate to perform feature reduction and simultaneous improve the Recognition Rate. Some practical examples prove that is possible to Recognise Portuguese Granites in 100%, with only 3 morphological features (from an original set of 117 features), which is well suited for real time applications. Moreover, the present method represents a robust strategy to understand the proper nature of the images treated, and their discriminant features. KEYWORDS: Feature Reduction, Genetic Algorithms, Nearest Neighbour Rule Classifiers (k-NNR).