Clustering, the task of grouping related items in a dataset into a number of groups, is a difficult common data mining task with a very large search space. The difficulty of clustering is increased further, when a dataset has many features (attributes): many clustering methods require large amounts of computational power and perform poorly at high dimensions. Clustering results utilising a large number of features are also difficult to interpret and understand, and often may incorrectly utilise on weaker or less-useful features. Feature reduction methods (feature selection and feature construction) have been shown to address these problems in classification tasks, but have had very little application to clustering problems. Evolutionary Computation, in particular, has been very successfully applied to clustering and feature reduction tasks, but has seen only very little use in applying feature reduction to clustering tasks. This thesis will aim to explore and introduce new approaches for improving the performance and interpret-ability of clustering by applying evolutionary computation to feature reduction on clustering problems. In particular, the use of Particle Swarm Optimisation for simultaneous clustering and feature selection, and Genetic Programming for feature construction in clustering problems will be explored. Focus is also placed on studying the design of fitness functions for these tasks, and on applying EC-based clustering to the closely related feature grouping problem.