The rexinoid bexarotene represses cyclin D1 by causing its proteasomal degradation. The epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) erlotinib represses cyclin D1 via different mechanisms. We conducted a preclinical study and 2 clinical/translational trials (a window-of-opportunity and phase II) of bexarotene plus erlotinib. The combination repressed growth and cyclin D1 expression in cyclin-E- and KRAS/p53-driven transgenic lung cancer cells. The window-of-opportunity trial in early-stage non-small-cell lung cancer (NSCLC) patients (10 evaluable), including cases with KRAS mutations, repressed cyclin D1 (in tumor biopsies and buccal swabs) and induced necrosis and inflammatory responses. The phase II trial in heavily pretreated, advanced NSCLC patients (40 evaluable; a median of two prior relapses per patient (range, 0-5); 21% with prior EGFR-inhibitor therapy) produced three major clinical responses in patients with prolonged progression-free survival (583-, 665-, and 1,460-plus days). Median overall survival was 22 weeks. Hypertriglyceridemia was associated with an increased median overall survival (P = 0.001). Early PET (positron emission tomographic) response did not reliably predict clinical response. The combination was generally well tolerated, with toxicities similar to those of the single agents. In conclusion, bexarotene plus erlotinib was active in KRAS-driven lung cancer cells, was biologically active in early-stage mutant KRAS NSCLC, and was clinically active in advanced, chemotherapy-refractory mutant KRAS tumors in this study and previous trials. Additional lung cancer therapy or prevention trials with this oral regimen are warranted.