This book focuses on some related problems in medical diagnosis for the identification of the perfect biomarkers which in turn help to classify the subjects into one of the two groups or populations. A brief introduction about the concept of clinical trials and a basic introduction to the concept of ROC curves have been discussed. Literature review has been presented on the three different approaches such as non-parametric, parametric and semi-parametric for estimating and testing various procedures of ROC curve. Specific spreadsheet templates and solutions have been developed to perform ROC curve analysis. These templates can be used for any type of dataset by simply posting the values in appropriate columns. A new method for evaluating the AUC has been proposed by considering the confidence intervals of means of both healthy and diseased populations. Identification of optimal cut point and the related biomarker has been made. Various cofactor combinations have been studied to provide the discriminant scores, in turn which can be used for further classification.