This paper proposes a clustering approach based on Modified Mutation strategy in the Differential Evolution (MMDE). Differential evolution is an evolutionary computation technique used for optimization. Though DE is very efficient, it sometimes suffers from the issue of slow convergence and the difficulty of achieving a global solution. To overcome these issues, in this paper, a modified mutation method was developed, which maintained the balance between exploration and exploitation. The objectives of modification were to achieve a higher rate of convergence and to obtain better cluster efficiency. The proposed form of modification had been applied on probabilistic environment to define the differential vector through randomly selected members and to obtain the best solution. Over the number of benchmark dataset, clustering efficiency had been estimated and compared with Conventional Differential Evolution (CDE) as well as Particle Swarm Optimization. The proposed method had been tested on a number of benchmark datasets. Experimental results had shown that MMDE had better and consistent clustering efficiency when compared to Conventional Differential Evolution (CDE) and Dynamic Weighted Particle Swarm Optimization (DWPSO).