Protein tyrosine kinases are enzymes which catalyze the phosphorylation of tyrosine residues and activate a downstream cascade of cellular signalling pathways which regulate cell proliferation, differentiation and apoptosis and a wide variety of cellular functions. Clinical developments over the past decade have identified several novel therapeutic agents which inhibit tyrosine kinase activity, either by direct receptor inhibition or indirect inhibition of tyrosine kinase controlled pathways. Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKI), such as gefitinib and erlotinib have been studied extensively in patients with refractory non-small cell lung cancer (NSCLC). Early studies with gefitinib showed undoubted clinical activity but failed to show a survival benefit, whereas studies with erlotinib showed a small but statistically significant benefit in overall survival. Subsequent studies explored the possibility of synergistic activity between targeted agents (gefitinib or erlotinib) and conventional chemotherapy drugs reporting disappointing results. Clinical trial results with gefitinib and erlotinib, either as monotherapy or in combination with chemotherapy, have failed to match the encouraging results noted in the pre-clinical setting. It is now increasingly recognised that clinical exploration of molecular targeted agents may not conform well to traditional phase I/II/III drug trial designs. Therapeutic responses may be limited to a small subpopulation of patients, therefore diluting the overall therapeutic effect. Hypothesising a genetic basis for the heterogeneity in trial results, biomarkers (such as EGFR gene mutation analysis, EGFR protein expression, and increased EGFR gene copy number) have been studied with a view to identifying a target population most likely to benefit from these drugs. Future clinical trials with targeted agents need to be carefully designed to incorporate correlative translational research elements that will allow selection of appropriate treatment strategies for individual patients. For assessment of phase III trial results in advanced disease, progression free survival may serve as a more appropriate end-point than response rate in an adequately designed trial in the appropriately selected population, although there should be no substitute for the overall survival and quality of life end points. The role of EFGR TKI in NSCLC will be discussed in detail and data from these studies will be used to illustrate the challenges in designing clinical trials and interpreting outcomes.