Background Dysregulated immunity is a hallmark of SARS-CoV-2 infection. Immune suppression is indicated by low monocyte expression of human leukocyte antigen D-related (mHLA-DR). T cells are important antiviral cells. We aimed to assess the role of mHLA-DR and T lymphocyte frequency in predicting COVID-19 severity. Patients and Methods This cross-sectional study enrolled 97 SARS-CoV-2 positive patients, including mild to moderate (n = 49) and severe cases admitted to intensive care unit (ICU) (n = 48). These ICU cases were further subdivided into survivors (n = 35) and non-survivors (n = 13). Results Severe cases had a significant decrease in the mHLA-DR mean fluorescence intensity (MFI) and T lymphocyte percentage compared to mild to moderate cases ( P <0.001). Non-survivors had a lower T lymphocyte percentage ( P =0.004) than survivors. The mHLA-DR MFI and T lymphocyte percentage correlated with oxygen saturation (r=0.632, P <0.001) and (r=0.669, P <0.001), respectively. According to the ROC curves, mHLA-DR MFI, at a cutoff of 143 and an AUC of 0.9, is a reliable biomarker for distinguishing severe COVID-19 cases, with 89.6% sensitivity and 81.6% specificity, while T lymphocyte frequency had 81.3% sensitivity and 81.6% specificity at a cutoff of 54.4% and an AUC of 0.9. The T lymphocyte percentage as a predictor of ICU survival at a cutoff of 38.995% exhibited 100% sensitivity and 57.1% specificity. According to multivariate regression analysis, reduced mHLA-DR MFI and T lymphocyte percentage are independent predictors of COVID-19 severity (OR = 0.976, 95% CI: 0.955–0.997, P = 0.025) and (OR = 0.849, 95% CI: 0.741–0.972, P = 0.018), respectively. Conclusion Reduced mHLA-DR expression and T-lymphocyte percentage are independent predictors of COVID-19 severity. Oxygen saturation percentage is correlated with mHLA-DR MFI and T lymphocyte frequency. The T lymphocyte frequency is a proposed predictor of COVID-19 survival in ICU admitted patients.