We investigated the associations of type 2 diabetes (T2D) and its related-traits, and their genetic susceptibility with the risk of severe COVID-19 and the mortality in COVID-19 patients. We analyzed the population-based cohort data of 460,238 individuals from UK Biobank to study the COVID-19 infection in T2D patients. First, we conducted unadjusted and adjusted logistic regression models to discover the association between severe COVID-19 and T2D, T2D-related traits such as BMI, waist circumference and waist-hip ratio. In the multivariable logistic models, we adjusted for potential confounders including age, sex and ethnicity. Second, we examined the association of overall genetic susceptibility for T2D with the severe COVID-19 using logistic models. To capture overall genetic predisposition for T2D, we computed polygenic risk scores (PRS), which represent an individual’s overall genetic risk for T2D trait, based on GWAS summary statistics for T2D in UK Biobank data. Third, we performed multivariable Cox proportional hazard models to investigate the effect of T2D patients infected with severe COVID19 on the length of survival. Based on the analyses, we found that individuals with T2D and those with more severe obesity are at a higher risk for developing severe COVID-19. We also found a positive association of the individual’s overall genetic risk for T2D and T2D-related traits with severe COVID-19, indicating the roles of T2D and T2D-related genetics in the pathogenesis of severe COVID-19. Furthermore, we found that patients with COVID-19 and T2D were associated with a significant decrease in mortality.