Background: Despite advances in available treatment, gastric cancer (GC) survival remains low (34.5% overall 5-year survival). In patients diagnosed with cancer, having a prior psychiatric disorder (PD) has been associated with worse survival outcomes compared with no prior PD. Therefore, we examined the relationship between a prior PD diagnosis and GC survival. Methods: We queried the Surveillance, Epidemiology and End Results (SEER)-Medicare database from 2000 to 2017 for patients who were 66 years of age or older with histologically confirmed GC. Patients with prior cancer history, who were diagnosed by autopsy or death certificate, or were not continuously enrolled in Medicare for at least one year prior to their GC diagnosis were excluded. Patients diagnosed with a PD at least one year prior to GC diagnosis and patients without a prior PD were defined as the exposed and unexposed groups, respectively. PDs were categorized as either mild (depression and anxiety) or severe (bipolar or psychotic disorders and schizophrenia). We performed adjusted multivariable Cox proportional hazards regression modeling to evaluate the relationship between prior PD and 5-year overall and cancer-specific survival. Cancer stage, histologic grade, age at diagnosis, sex, race and ethnicity, rural/urban, comorbidity, socioeconomic status, and marital status were examined as potential confounders. Results: Out of 10,464 identified GC patients, 1,651 had a prior PD (46.3% male, mean age 77.5 years) and 8,813 did not have a prior PD (61.5% male, mean age 77.4 years). Our PD patient cohort included 1,355 mild and 296 severe PD patients. When we examined overall survival, patients with a prior PD had a greater hazard of death compared to patients without a prior PD (adjusted hazard ratio: 1.12; 95% CI: [1.05, 1.19]). However, we found no association between having a prior PD compared with not having a prior PD on GC-specific survival (adjusted hazard ratio: 1.07; 95% CI: [0.99, 1.15]). In GC patients with a prior PD, we examined the effects of severe PD versus mild PD on survival. For overall survival, severe PD patients had a greater hazard of death compared with mild PD patients (adjusted hazard ratio: 1.43; 95% CI: [1.23, 1.65]). Similarly, for GC-specific survival, severe PD patients had a greater hazard of death compared to mild PD patients (adjusted hazard ratio: 1.26; 95% CI: [1.06, 1.50]). Conclusions: Our study found that GC patients with a prior PD had a 12% greater likelihood of overall mortality compared with GC patients without a prior PD. Our analysis also showed that GC patients with a prior severe PD had a 43% higher likelihood of overall mortality and a 26% higher likelihood of GC-specific mortality compared with GC patients with a prior mild PD. Future work to identify the mechanism between PDs and cancer survival is critical to develop effective targeted interventions to improve survival of this highly fatal cancer.