Chemometric processing of laboratory results creates new possibilities for supporting decisions on the right diagnosis in laboratory medicine and clinical chemistry. The approach, originally applied in medicine, has recently been implemented also in analytical chemistry. Qualitative characteristics of laboratory methods, like sensitivity, specificity or efficiency are used for the selection of an appropriate decision limit, according to which all examined objects or subjects can be classified into two categories (e.g., healthy and ill subjects). Assessment of performance of the compared laboratory methods is made using ROC (receiver operating characteristic) curves. Utilization of multivariate statistical methods, combining original variables in an optimum way, increases the diagnostic value of laboratory assays. A case study based on the multicomponent calculation by principal component analysis shows how a linear combination in the analysed tumor markers increases their diagnostic value for the differential diagnostics of pleural malignancy.