Cytosolic and organelle-based heat-shock protein (HSP) chaperones ensure proper folding and function of nascent and injured polypeptides to support cell growth. Under conditions of cellular stress, including oncogenic transformation, proteostasis components maintain homeostasis and prevent apoptosis. Although this cancerrelevant function has provided a rationale for therapeutically targeting proteostasis regulators (e.g., HSP90), cancer-subtype dependencies upon particular proteostasis components are relatively undefined. Here, we show that human rhabdomyosarcoma (RMS) cells, but not several other cancer cell types, depend upon heatshock protein 70 kDA (HSP70) for survival. HSP70-targeted therapy (but not chemotherapeutic agents) promoted apoptosis in RMS cells by triggering an unfolded protein response (UPR) that induced PRKR-like endoplasmic reticulum kinase (PERK)-eukaryotic translation initiation factor α (eIF2α)-CEBP homologous protein (CHOP) signaling and CHOP-mediated cell death. Intriguingly, inhibition of only cytosolic HSP70 induced the UPR, suggesting that the essential activity of HSP70 in RMS cells lies at the endoplasmic reticulumcytosol interface. We also found that increased CHOP mRNA in clinical specimens was a biomarker for poor outcomes in chemotherapy-treated RMS patients. The data suggest that, like human epidermal growth factor receptor 2 (HER2) amplification in breast cancer, increased CHOP in RMS is a biomarker of decreased response to chemotherapy but enhanced response to targeted therapy. Our findings identify the cytosolic HSP70-UPR axis as an unexpected regulator of RMS pathogenesis, revealing HSP70-targeted therapy as a promising strategy to engage CHOP-mediated apoptosis and improve RMS treatment. Our study highlights the utility of dissecting cancer subtype-specific dependencies on proteostasis networks to uncover unanticipated cancer vulnerabilities.T he synthesis and folding of proteins is a highly regulated process in both normal and malignant cells (1). The protein homeostasis ("proteostasis") network that operates in cancer cells offers targets for therapeutic development, such as the proteasome and specific protein chaperones (2, 3). For example, the heat-shock protein 90 kDa (HSP90) chaperone maintains the stability of some mutant oncoproteins and permits the outgrowth of drug-resistant cells, fueling the development of small molecule HSP90 inhibitors as anticancer agents (4). Despite promising early results, these HSP90 inhibitors do not show large-scale clinical success across the majority of cancers tested (4). Proteasome inhibitor treatment has made a substantial impact on cancer patient outcomes but to date has been highly effective in only a small number of malignancies (5). Although the proteostasis network provides an increasingly rich landscape beyond these two targets, the dependence of particular cancer subtypes on specific proteostasis components is not well defined. Filling this knowledge gap is essential to elaborate the role of proteostasis in the pathogenesis...