Background: The association between obstructive sleep apnea (OSA) and the incidence and mortality of cancer remain unclear, especially in Asian populations. Thus, this study was conducted to explore the relationship between OSA and the incidence and mortality of cancer in hospitalized patients. Methods: This retrospective cohort study evaluated inpatients from Guangdong Provincial People's Hospital for suspected OSA between January 2005 and December 2015. Cancer incidence, all-cause mortality, and cancer mortality and were determined using data from the hospital information system and Centers for Disease Control. Between-group comparisons were carried out by performing a chi-square test and analysis of variance. Kaplan-Meier analysis and the Cox proportional risk model were applied to investigate the association between OSA and cancer incidence and mortality. Results: Of the 4,623 hospitalized patients included, 3,786 (81.9%) patients were diagnosed with OSA. After a median follow-up of 9.1 years (interquartile range, 9.79-11.44), the incidence of cancer was 6.6% (251/3,786), with lung cancer having the highest incidence at 1.6% (60/3,786). The mortality rate of OSA patients was higher than that of non-OSA patients (16.83% vs.12.78%, p=0.008), but the relationship between apnea-hypopnea index (AHI), oxygen saturation less than 90% (TSat90), and cancer mortality was not statistically significant (p>0.05).The mortality rate for all types of cancer was 2.8% (105/3,786), with lung cancer having the highest mortality rate at 0.8% (32/3,786). The cumulative incidence of cancer in the severe OSA group was 8.2%, which was higher than that in the normal, mild, and moderate OSA groups (P=0.010). Further, the Cox proportional risk regression model showed a progressive enhancement in the risk of cancer incidence as the AHI increased (adjusted hazard ratio [HR]: 1.009 [95% confidence interval (CI): 1.003-1.016], P=0.005). Based on subgroup analysis, the risk of cancer increased as the AHI increased in patients aged <65 years (adjusted HR: 1.019 [95% CI: 1.007-1.031], P=0.002). In addition, the cancer incidence was significantly higher in the severe OSA group than in the normal, mild, and moderate OSA groups (adjusted HR: 2.825 [95% CI: 1.358-5.878], P=0.019). Conclusion: The incidence of cancer is higher in patients with OSA than in non-OSA patients and is significantly positively associated with the severity of OSA. Particularly, for OSA patients aged <65 years, lung cancer is the main cause of death in those with new-onset cancer. Mortality was higher in OSA patients than in non-OSA patients.