The impact of KRAS mutations on the efficacy of therapies that target the epidermal growth factor receptor (EGFR) is a major, ongoing area of oncology research, aimed at identifying the best possible treatments for individual colon cancer patients. Because patients with KRAS mutant colorectal tumors rarely respond to anti-EGFR monoclonal antibodies, testing is required to confirm the patient's tumor is KRAS wild-type before utilizing these therapies. Despite being studied for more than 30 years, new information continues to develop regarding KRAS and its role in colon carcinogenesis. This information must be integrated into the development of effective colon cancer treatment strategies. This review will summarize recent evidence that most, if not all, colon tumors encompass at least a subpopulation of KRAS mutant cells, meaning tumors characterized as KRAS wild-type are in most cases tumors with relatively low KRAS mutant tumor cell content. Recent studies support the hypothesis that relapse in advanced colorectal patients treated with EGFR-targeted monoclonal antibody therapy involves the outgrowth of previously undetected KRAS mutant tumor cell populations. Studies investigating the effects of oxidative stress on Ras signaling suggest that the frequent presence of minor KRAS mutant tumor cell populations may be a consequence of hypoxic conditions within tumors, which produce a negative selection against KRAS mutant cells in polyclonal tumors. Thus, the literature and current practices for characterizing tumor KRAS mutation don't accurately reflect the nature of colon tumor KRAS mutation, even though an accurate understanding is critical for identifying the best strategies for intervention.