Lung adenocarcinoma (LUAD) and pancreatic ductal adenocarcinoma (PDAC) together account for ~15% of all cancer mortalities. Both LUAD and PDAC are commonly driven by KRAS mutations. The first allele-specific KRAS inhibitors were recently approved for LUAD, but clinical benefit is limited by intrinsic and acquired resistance. LUAD predominantly arises from alveolar type 2 (AT2) cells, which function as facultative alveolar stem cells by self-renewing and replacing alveolar type 1 (AT1) cells. Using genetically engineered mouse models, patient-derived xenografts and patient samples, we found inhibition of KRAS promotes transition to a quiescent AT1-like cancer cell state in LUAD tumors. Similarly, suppressing KRAS-induced AT1 differentiation of wild-type AT2 cells upon lung injury. The AT1-like LUAD cells exhibited high growth and differentiation potential upon treatment cessation, whereas ablation of the AT1-like cells robustly improved treatment response to KRAS inhibitors. Intra-tumoral heterogeneity in PDAC is characterized by a balance between basal and classical epithelial cancer cell states, with basal dominance associated with chemoresistance and a dismal prognosis. Using genetically engineered mouse models and patient-derived xenografts, we found that basal PDAC cells are highly sensitive to KRAS inhibitors. Employing fluorescent and bioluminescent reporter systems, we longitudinally tracked cell-state dynamics in vivo and revealed a rapid KRAS inhibitor-induced enrichment of the classical state. Lineage tracing identified these enriched classical PDAC cells as reservoirs for disease relapse. Genetic ablation of the classical cell state is synergistic with KRAS inhibition, providing a pre-clinical proof-of-concept for this strategy. Our results uncover an unexpected role of KRAS in promoting intra-tumoral heterogeneity and suggest targeting alveolar or classical epithelial differentiation may augment KRAS-targeted therapies in LUAD and PDAC, respectively.