Emerging results from clinical trials of epidermal growth factor receptor (EGFR) inhibitors indicate good tolerability and at best modest to no clinical activity in a variety of human carcinomas. The lack of molecular evidence to suggest a pathogenic role for the EGFR in most common solid tumors is not inconsistent with the modest clinical activity observed in these single-agent trials. In addition, the lack of enrollment of patients with EGFR-dependent tumors onto these trials leaves open the possibility that these studies may have included some tumors that benefited from these therapies but for which this benefit was diluted and thus not detected within the enrolled patient population at large. These considerations suggest that the enrollment of unselected patients onto trials with EGFR inhibitors should be revisited to identify a cohort of EGFR-dependent tumors against which EGFR inhibitors might exhibit clinical activity. This approach would also exclude patients for whom these drugs will not induce any clinical benefit.