Patients suffering from severe respiratory symptoms as a result of SARS-CoV-2 infection can rapidly generate virus-attacking T cells, and can increase this production over time, suggests a new study of T cells from 10 COVID-19 patients under intensive care treatment. In addition, 2 out of 10 healthy individuals without prior exposure to the virus harbored SARS-CoV-2-reactive T cells, the researchers found, possibly indicating that these T cells can cross-react to the novel coronavirus due to past infection with related coronaviruses that cause common cold symptoms. Together, these new data address the poorly understood question of whether SARS-CoV-2-specific T cell responses vary in patients over time depending on disease severity, and helps to answer whether patients with more severe symptoms can generate protective virus-specific T cells at all. The study also provides new clues regarding the cells responsible for excessive immune responses, including life-threatening "cytokine storms," and may also help inform vaccine design. Daniela Weiskopf and colleagues extracted blood cells from 10 patients at weekly intervals starting soon after they were admitted to the ICU for COVID-19 and exposed these cells to "megapools" of known SARS-CoV-2 epitopes - a technique meant to capture a large fraction of total viral-reactive T cells. They found SARS-CoV-2-specific CD4+ helper T cells in all 10 patients and CD8+ "killer" T cells in 8 out of 10 patients, and characterized the cells' production of specific inflammation-triggering cytokines. The strongest T cell responses were directed to the virus' spike (S) surface glycoprotein, supporting prior work that has pointed to the S protein as a promising target to induce virus-specific T cells. Furthermore, screening all patients at 0, 7, and 14 days after inclusion in the study revealed that SARS-CoV-2-specific T cells were present relatively early during the course of infection and increased in these patients over time. Using the same T cell stimulation technique in age-matched healthy controls, the researchers found SARS-CoV-2-reactive T cells in 2 out of the 10 individuals. Based on their findings, the authors note promising areas for future work, including an investigation of how preexisting SARS-CoV-2-specific T cells in healthy controls correlate to protection against COVID-19 disease, as well as identification of T cell types responsible for cytokine storms.