Every pathology test starts off with an optimistic future of being able to distinguish health and disease. This is followed by disillusionment when we find out that the marker may be absent in disease and present in health. Because of the prognostic implications of malignancy, these hopes and fears become exaggerated for tumour markers. Tumour markers are seldom proteins specifically made by tumours and are usually normal products whose physiological role and determinants need to be understood before we can understand how to interpret their levels in disease. The problematic issue of the role of PSA in prostate cancer has led to the situation today where I believe we understand the PSA test at least as well as any other test in the pathology laboratory. The many PSA issues of physiological determinants, multiple forms, assay standardisation and variability, rate of rise, value in screening, monitoring and prognosis serve as templates for the better understanding of all tumour markers, and many other tests. Like any laboratory test, the indiscriminate use of tumour markers can lead to low predictive value where they cause more harm than good. Once the strengths and weaknesses of tumour markers are understood, they add value to clinical decision making. If the strengths and weaknesses aren’t understood, we may be better off without them. There is a lot we have learnt about all the tumour markers including CEA, CA125, aFP, CA19.9, CA15.3, aFP, but it seems as if the anxiety and emotion surrounding their use is a barrier to applying this knowledge for a more worthwhile application.