Helper T cells (CD3+CD4+ T cells) and cytotoxic T cells (CD3+CD8+ T cells) play direct and indirect antiviral roles. This study retrospectively explored the clinical significance of peripheral lymphocytes, especially the dynamic analysis of T-cell subsets, in determining coronavirus disease 2019 (COVID-19) severity and progression. Seventy-nine patients with COVID-19 in the Public Health Clinical Center of Chengdu from January to February 2020 were included, 59 of which were analyzed for dynamic peripheral T-cell subsets expression. The neutrophil to CD4+ T lymphocyte ratio (N4R) and neutrophil to CD3+ T lymphocyte ratio (N3R) showed clinical significance in differentiating severe or critically-severe COVID-19, with area under receiver operating characteristic curves (AUCs) of 0.933 and 0.900, respectively (P < 0.05). COVID-19 patients with more baseline peripheral lymphocytes or NK cells were prone to test negative to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) after therapy (P < 0.05), and the AUC of NK cells for predicting negative results of SARS-CoV-2 RNA detection after therapy was 0.800. When the number of peripheral CD3+CD4+ and CD3+CD8+ T cells in COVID-19 patients continuously increased 6-9 days after baseline, the period of disease exacerbation could be delayed for more than 2 weeks after admission. Baseline N4R and N3R could be potential biomarkers for assisting in differentiating COVID-19 severity, and dynamically monitoring peripheral CD3+CD4+ and CD3+CD8+ T cells 6-9 days after baseline could help clinicians to evaluate disease progression in COVID-19 patients.