Essentially, all of our cells express surface molecules (HLA) that perform a kind of "real time analysis of the functional genome": Samples of peptides representing all gene products expressed by a cell are presented by HLA molecules. This allows our T cells to sense the integrity of all cells. Alterations, including mutations intrinsic to tumour cells, can thus be sensed by T cells. Each individual tumour expresses its own set of mutations (an average of 90 mutations in solid tumours have been identified, for example by the Vogelstein lab), some being essential for tumour growth, others being passenger mutations. Both categories can give rise to immunogenic peptides recognizable by T cells that can kill the tumour cells. In addition, tumour cells show overexpression of several gene products which can also give rise to correspondingly higher density of the respective HLA-presented peptides that again can be sensed by T cells. Thus, we are confronted with two layers of individuality: Each human being has his or her own set of HLA molecules, with individual peptide specificities, and each tumour owns its individually accumulated set of mutations and overexpression of genes. The ultimate consequence for immunotherapy of cancer is thus: Identify the mutations/alterations in the tumour, identify the new peptides presented on the tumour's HLA, synthesize these peptides, individually for each patient, and vaccinate the very same patient with these peptides, using appropriate adjuvants and conditions. The feasibility as well as the challenge of this strategy will be discussed.