Crime prediction and analysis is not an easy approach for government officials. Several prediction systems have been developed to assist the government agencies in solving crimes but most of them lacked essential components which would have made the quality of their output accurate, reliable and quick. This paper therefore aims to apply clustering techniques that has been developed for Criminal Prediction. The aim is to propose a better clustering technique for criminal prediction in order to provide investigators with rich sources of intelligence and information which can be used to predict and prevent criminal activity. The main goal is to provide accurate results quickly which can help government officials to predict criminal and crime prone areas.