Regarding the evolution of the treatments of colorectal cancer during the last five years, it appears that numerous questions have to be considered with a strategic point of view. In order to avoid inclusion of a lot of patients in clinical patients, we urgently need the help of biology to give us arguments to choose one treatment or another. The fact that ten years after their approval, we are not able to select responders to oxaliplatin or irinotecan, confirm that this is difficult problem to solve. The contribution of biology to the prescription of drugs commonly used in colorectal cancer is discussed in this paper. There are already situations where the contribution of biology to clinical practice and prescription is not debatable : this is the case for the use of UGT1A1 status determination when using irinotecan and the determination of KRAS status for the prescription of panitumumab (and cetuximab in a few months). This individual adaptation is a dream that becomes reasonable when we look on the recent results concerning EGFR inhibitors, but a lot of work has to be done and it is not sure that biological assessment of the tumour will be able to solve all the problems. For instance, to determine predictive factors of response to angiogenesis inhibitors, it is likely that solutions will come from new techniques of imaging rather than from biology. However, new tools such as proteomics or metabolomics, as well as a better dialogue between clinician and biologist, will allow fast improvements. It must be emphasised that, for the first time in 2008, it is possible to prescribe targeted therapies to a specific "targeted" group of patients with metastatic colorectal cancer.