Mounting evidences have indicated the effects of the tumor micro-environment (TME) on gastric cancer (GC) prognosis, tumorigenesis and metastasis. However, there is no reliable biomarker that could effectively predict the prognosis and estimate the immune feature of GC patients. Herein, we obtained the expression data from The Cancer Genome Atlas (TCGA) and the National Center for Biotechnology Information (NCBI) databases to construct a deacetylation regulator associated risk signature that could be used as a predictor of survival outcome and immune feature of GC patients. A total of 13 genes were screened out from the univariate-LASSO-multivariate analyses to generate the risk signature based on the regression coefficients. The risk signature exhibited robust prognostic abilities in both training set and testing set, and this association remained significant in the multivariate analysis after controlling for age, gender, T, N, M or Stage. Functional analyses indicated that the risk signature showed a correlation with immune feature of GC patients. Further analyses indicated the risk signature could be used to estimate the infiltration immune cells and predict the immunotherapy response of GC. Taken together, our risk signature could predict prognosis and immune feature of CC, which might provide crucial clues to formulate individualized treatment and provides promising novel biomarkers for immunotherapy.