This paper introduces a multi-objective EA, termed the Clustering Pareto Evolutionary Algorithm (CPEA). The CPEA finds and retains many local Pareto- optimal fronts, rather than just the global front as is the case of most multi- objective EAs found in the literature. This has been achieved using a clustering technique commonly used in multivariate statistical analysis, which ensures that competition between individuals is local in variable space, allowing the population to grow to resolve as many Pareto-optimal fronts as necessary. The performance of the CPEA is evaluated on several test problems taken from the literature which have either single optima or multiple local optima and is shown to be extremely effective. The present clustering method is computationally expensive and will be replaced with an incremental method in the near future.