Fuzzy clustering method is preferred by researchers in image segmentation for digital images. The choice of distance metric is highly important for clustering process. Results of segmentation in case of fuzzy clustering entirely depend on the distance metric used for the clustering process. This paper evaluates the performance of Fuzzy Clustering process for noisy image segmentation using various distance metrics namely Euclidean, Manhattan, Minkowski, Kernel, Kernel RBF, New distance, Canberra, Pearson, Chebyshev and Eisen Cosine. The performance of distance metrics in the clustering approach is tested with two digital images. The performance is quantitatively accessed using five metrics namely Partition Entropy, Partition Coefficient, Fukuyama-Sugeno, XieBeni function and Tanimoto index. After a thorough qualitative and quantitative examination of the segmentation results, it is discovered that in maximum cases, specifically in the presence of noisy images, the Manhattan distance metric outperformed every other distance metric.