Cumulative evidence suggests that long non-coding RNAs (lncRNAs) may be good biomarkers in various types of tumors. In the present study, we mined lncRNA expression profiling in 739 lung cancer patients from Gene Expression Omnibus (GEO) datasets. A risk score model was constructed based on the expression data of these eight lncRNAs in the training dataset (GSE30219). The validation for the association was performed in three independent testing sets (GSE31210, GSE37745 and GSE19188). Finally, a set of eight lncRNA genes (AK021595, BC030759, AK000053, AK124307, BC020384, AK022024, CR615992 and AF085995) were identified by the random survival forest algorithm. Using a risk score based on the expression signature of these lncRNAs, we separated the patients into low-risk and high-risk groups with significantly different survival times in the training set. This finding was validated in the other three testing sets. Further study revealed that the eight-lncRNA expression signature was independent of age and gender. Gene Set Enrichment Analysis (GSEA) suggested that lncRNAs were involved in cell cycle and DNA replication signaling pathways. Therefore, the eight lncRNAs may be candidate prognostic biomarkers for lung cancer patients.