The number of potential molecular markers is constantly increasing. However, for most malignancies there is no simple test to detect early-stage tumours that is useful for screening purposes because most biomarkers have poor sensitivity or specificity, or other clinical value. One approach to increase their value is to measure several biomarkers at a time. The additional information should always yield a test more able to distinguish between patients and healthy individuals, and ideally between different kinds of tumour. Our work in colorectal, lung, and head and neck cancer, illustrates the evolution of this idea. A test in which ELISAs for key serum markers are arrayed based on immunoblot technology or flow cytometric beads is suggested because these techniques are more transferable to practical application in clinical decision making. Moreover, the multivariate data obtained from such a test can easily be managed with statistical methods already developed in the fields of genomics and proteomics.