Abstract: Class discovery is one of the most important tasks in cancer classification using biomolecular data. To perform this, a multiple clustering approach called Hierarchical clustering is used. It uses one of the metrics called Manhattan Distance which measures the distance between the values of the data set and builds a hierarchy of clusters after analysing it. The clustering result enables to classify the cancer types and it is further evaluated by Range check and Delta check. The various test results are compared with the known initial range of values using Range check. Delta check is performed on the current test result and the immediate previous test result for better results. These techniques are used to improve the diagnosis of cancer.