Clustering is an important mission in the field of machine learning, pattern recognition and web mining. Handling uncertain data in the information system is one of the key research topics in the vicinity of knowledge representation. Number of clustering algorithms are available [23][6][12]27]; but many of those algorithms are challenging when dealing with uncertain data. The aim of the paper is to tune two existing rough c-means and fuzzy c-means and integrate them into a tuned hybrid soft clustering algorithm termed as the tuned rough-fuzzy c-means algorithm. Rough c-means is extremely sensitive to the initial placement of the cluster centers. The proposed algorithm is enhanced by introducing dynamic centroid computation. The proposed algorithm performance is compared with the existing rough c means, fuzzy c-means, and rough fuzzy c –means approaches. The effectiveness of the algorithm is verified on real and synthetic datasets. © 2015, Institute of Integrative Omics and Applied Biotechnology. All rights reserved.