Abstract Background Endoplasmic-reticulum-stress (ER-Stress) chaperones like the main ER-Stress moderator GRP78 (glucose-regulated-protein, 78kDa) are involved in the pathogenesis of coronary artery disease (CAD). In addition to their established intracellular localization, chaperones are also secreted into the extracellular space. However, the significance of extracellular ER chaperones in CAD remains unclear. Here, we investigated the role of extracellular GRP78 in patients with suspected CAD and its impact on endothelial cell inflammation. Methods Serum concentration of GRP78 was measured by ELISA in 622 patients undergoing coronary angiography (70.0 ± 12.2 years, 33.6% female and 28.4% with acute coronary syndrome: ACS). The primary endpoint was defined as one-year mortality. Human coronary artery endothelial cells (HCAEC) were treated with ER-Stressors and the effects of ER-Stress conditioned medium (CM) on recipient cell viability, apoptosis, reactive oxygen species (ROS) formation and gene expression were examined. Results GRP78 levels were higher in blood samples of patients with CAD than in patients without CAD (2584 ng/ml vs. 2061 ng/ml, p = 0.047). In patients presenting with ACS, GRP78 levels were lower than in patients presenting with chronic coronary syndrome. Elevated GRP78 levels were independently associated with lower one-year mortality, even after adjustment for cardiovascular risk factors and demographic parameters (4.5% vs. 9.0%, log-rank p = 0.022; Fig.1A). Increased levels of GRP78 (> median of 1710 ng/ml) were associated with a higher BMI (29.0 kg/m2 ± 7.3 vs. 27.0 kg/m2 ± 6.2, p < 0.0001) while there was no difference regarding age, sex, or cardiovascular risk factors. Mass spectrometry analyses of the HCAEC secretome and ELISA revealed increased extracellular GRP78 secretion after ER-Stress activation. Co-incubation with Brefeldin A, an inhibitor of ER-Golgi-trafficking, confirmed active regulation of GRP78 secretion. GRP78-containing CM improved viability while reducing apoptosis, formation of ROS, ER-Stress-signaling, and inflammation in recipient HCAEC compared to CM lacking GRP78 due to BFA treatment or GRP78-knockdown. Likewise, recombinant GRP78 improved HCAEC viability while reducing expression of markers of inflammation and ER stress (Fig.1B). CRIPTO is a membrane-bound co-receptor for growth factors and was previously linked to GRP78-signalling. siRNA-Knockdown of CRIPTO reduces viability and leads to cell apoptosis, irrespective of GRP78-treatment (relative viability: Scr-siRNA+1µg/ml GRP78 vs. Criptodown+1µg/ml GRP78 1.22 ± 0.23 vs. 1.02 ± 0.13, p<0.01, Fig.1C). Thus, GRP78-mediated effects may be mediated by CRIPTO. Conclusion GRP78 is significantly increased in patients with CAD. Intriguingly, elevated levels are associated with reduced one-year mortality. ER-Stress-induced GRP78 secretion represents a feedback mechanism to ameliorate ER-Stress and inflammation in recipient endothelial cells (Fig.2).