Abstract KRAS mutations drive resistance to diverse targeted therapies, most notably to EGFR inhibitors in the setting of metastatic colorectal cancer (CRC). Curiously however, through a series of genetic screens involving constitutive activators of oncogenic signaling pathways, we found that ectopic expression of mutant HRAS, another RAS gene not typically associated with resistance, drove substantially stronger therapeutic resistance than mutant KRAS. Although HRAS and KRAS share ~85% amino acid sequence identity, the nucleotide sequence between the two varies enormously, with HRAS being enriched in common codons that yield high protein expression, while KRAS is enriched in rare codons that yield poor expression. This suggested that rare codons may limit the ability of KRAS to impart resistance in the clinic. Consistent with this notion, we find that primary resistance to the EGFR inhibitor cetuximab in CRC is dependent not only upon KRAS mutational status, but also upon the ability of cancer cells to overcome the translational barrier imposed by codon bias. Similarly, we show that more potent KRASQ61 mutations drive acquired resistance even in the setting of low protein expression, perhaps explaining the paradoxical enrichment of these mutations observed in patients with cetuximab-refractory CRC. Finally, we demonstrate that cancer cells globally upregulate translation in the setting of KRASG12-driven acquired cetuximab resistance, resulting in hypersensitivity to diverse small molecule inhibitors of translation. These findings demonstrate that codon bias plays a critical regulatory role in KRAS-driven therapeutic resistance and provide a mechanistic rationale for targeting protein translation to overcome resistance.