Discovery of new serum biomarkers exhibiting an increased sensitivity and specificity for cancer are of major importance for diagnosis improvement. There is considerable evidence for an immune response to cancer in humans, as demonstrated in part by the identification of autoantibodies against a number of tumour-associated antigens in sera from patients with different cancer types. Thus, identification of tumour-associated antigens and their associated antibodies is a promising strategy to find relevant biomarkers. Proteomic approaches such as SEREX and SERPA have allowed identification of great numbers of antigens and their cognate autoantibodies during these past few years. They show many advantages, and allowed identification of relevant autoantigens in different types of cancer. However, they are also time consuming, and lack sensibility and specificity. To circumvent these drawbacks, new proteomic techniques, based on protein or antibody arrays, allow high throughput analysis of multiple targets in a single experiment. Specific combinations of markers should thus be identified, theoretically being more efficient to detect a tumor compared to a single marker. In conclusion, these approaches promise great advances in the field of biomarkers for cancer. They also further need to be validated for clinical application on large populations.