BACKGROUND: De novo T cell expansion to a novel viral infection is assumed to confer protection, but empirical evidence in humans is limited. The SARS-CoV-2 pandemic provided a unique opportunity to investigate de novo T cell-mediated protection in antigen-naïve individuals without the confounding effects of preexisting immune memory. METHODS: We leveraged a prospective household contact study to recruit new COVID-19 cases a median of 4 days post-SARS-CoV-2 exposure. We longitudinally enumerated SARS-CoV-2 antigen-specific functional T cell subsets using dual IFN-γ/IL-2 fluorescence-linked immunospot (FLISpot) assays. We then correlated T cell dynamics with detailed clinical and virological outcomes derived from longitudinal measurement of symptom burden and viral load. FINDINGS: Early expansion (day 0-7) of SARS-CoV-2-specific IFN-γ-secreting T cells correlated with lower peak viral load and symptom burden. Conversely, late T cell expansion (day 7-28) correlated with higher symptom burden. Neither pre-existing cross-reactive T cells nor early antibody induction correlated with virological outcomes. INTERPRETATION: These findings provide empiric evidence for early antigen-specific T cell expansion being protective against naturally acquired viral infection in humans. FUNDING: This work is supported by the NIHR Health Protection Research Unit in Respiratory Infections, Imperial College London in partnership with the UK Health Security Agency (Grant number: NIHR200927; AL) and the Medical Research Council (Grant number: MR/X004058/1).