Conventional chemotherapeutic regimens have limited impact against most solid tumors and deal with significant toxicity. Over the last 10 years, novel anticancer treatments targeting specific molecules or genes involved in cancer progression have been developed to improve outcome and reduce side effects. In particular, the tyrosine kinase inhibitors gefitinib and erlotinib have been approved for the treatment of non-small-cell lung cancer. Their clinical activity has been related to different clinical and biological parameters, such as EGFR-activating mutations. However, not all clinical outcomes, including tolerability, are explained, and the identification/validation of novel biomarkers is a viable area of research. Germline polymorphisms can be easily assessed in blood samples, and candidate polymorphisms in EGFR and ABCG2 have been correlated with outcome and toxicity in non-small-cell lung cancer patients treated with gefitinib or erlotinib. However, differences in study population and design resulted in several controversial findings, while the prognostic versus predictive role of the polymorphisms still needs to be validated within larger prospective studies. More studies on the relationship of the genotype with drug pharmacokinetics and mechanism of action are also warranted. These future studies, as well as further development and application of novel technologies to decipher genetic alterations, might contribute to the validation of selected polymorphisms as molecular markers predictive of drug activity and help in the selection of tyrosine kinase inhibitors best suited to the individual patient.