The most important change that will emerge over the coming decade for management of cancer is the shift from conventional chemotherapy and radiotherapy to novel targeted therapies. The rapid expansion in the understanding of the molecular biological basis of cancer provides potential targets for novel therapies. Many molecules, with a variety of cellular targets are now entering the clinic with the emergence of some promising data. The key now is to define the patient population most likely to benefit from these agents through identification of clinical and biological markers indicating a sensitive tumour phenotype. The ongoing clinical development of signal transduction inhibitors presents several challenges to the existing dogma of clinical trial design. For example, in early phase trials the traditional endpoint of objective response rate may not be the most useful in selecting cytostatic agents, which nevertheless may possess clinically relevant activity. Rather, such trials should focus on the in vivo measurement of biological activity in order to define the optimum schedule of treatment that results in maximal inhibition of the therapeutic target. In later phase trials, endpoints such as progression free survival may be more useful than response rate, although of course there is no substitute for the endpoints of overall survival and quality of life, improvements in which these agents must demonstrate for them to be accepted into the cancer treatment armamentarium. Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKI) that have entered clinical trials will be discussed and the successes and failures of these trials will be used to illustrate the obstacles that confront the assessment of the activity of these agents so that this may guide the future development of other new agents by optimising clinical trial design. Signal transduction inhibitors may allow us to overcome resistance or restore sensitivity to conventional chemotherapy. The integration of these molecules with existing therapies should be based on robust pre-clinical data indicating potentially beneficial additive or even synergistic interactions. The correct clinical management strategy can be guided by preclinical modelling but can only be validated by carefully designed clinical trials. These will at the very least need to be conducted with correlative translational research elements that will allow us to select the most appropriate treatment strategy for individual patients.