To identify a cancer stemness-related prognostic signature, we analyzed RNA-seq data of patients with lung adenocarcinoma from the TCGA and GEO cohort using univariate Cox and LASSO analysis. Eight genes (PLEK2, GPX8, TCN1, PERP, CD79A, PAX5, FSCN1, and CNNM1) were eventually identified to develop a cancer stemness-related signature. AUCs for 5-year overall survival were 0.681, 0.673, and 0.774 in the training, validation, and test sets, respectively. Kaplan-Meier curve analysis showed that the high-risk group has a poor survival as compared with the low-risk group in the training set, validation set, and test set, respectively (P < 0.05). Six genes (PLEK2, GPX8, TCN1, PERP, FSCN1, and CNNM1) were significantly upregulated in the cancerous tissue compared with in the normal tissue (P < 0.001), and correlated with an unwanted prognosis (P < 0.05). We constructed a prognostic nomogram combining the cancer stemness-related signature with TNM staging to precisely assess the prognosis of the patient with lung adenocarcinoma. Finally, we observed that the stemness-related gene signature also could predict the malignant levels of lung adenocarcinoma possibly via hypoxia, epithelial-mesenchymal transition and angiogenesis. Conclusively, this study developed a novel cancer stemness-related gene signature, and identified six pivotal genes which can serve as potential therapeutic targets in the treatment of lung cancer.