This paper deals with a new approach for image segmentation of brain image by applying k-means clustering algorithm for detection of tumor. Manual segmentation of brain tumor is time consuming and challenging task. In image segmentation, clustering algorithms are very popular as they are intuitive and implementation is easy. One of the most widely used algorithms in the literature is the K-means clustering algorithm. This paper proposes a color-based segmentation method that uses K-means clustering technique. In kmeans algorithm partition of an image is takes places into k clusters. The K-Means algorithm produces accurate segmentation results only when applied to images defined by homogenous regions with respect to color. At first, the pixels are clustered based on their color. After clustering of pixels clustered blocks are merged to a specific number of regions. This approach provides a useful new solution for image segmentation in which tumor is detected. At the end of the process the tumor is extracted from the MIR or CT scan image and tumor position determined. It clarify the effectiveness of our approach to improve the segmentation quality in aspects of precision and computational time.