Background. The novel coronavirus (SARS-CoV-2) has produced a considerable public health burden but the impact that contracting the disease has on mental health is unclear. In this observational population-based cohort study, we examined longitudinal changes in psychological distress associated with testing positive for coronavirus disease 2019 (COVID-19). Methods. Participants (N = 8002; observations = 139 035) were drawn from 23 waves of the Understanding America Study, a nationally representative probability-based online panel of American adults followed-up every 2 weeks from 1 April 2020 to 15 February 2021. Psychological distress was assessed using the standardized total score on the Patient Health Questionnaire-4. Results. Over the course of the study, 576 participants reported testing positive for COVID-19. Using regression analysis including individual and time-fixed effects we found that psychological distress increased by 0.29 standard deviations ( p < 0.001) during the 2-week period when participants first tested positive for COVID-19. Distress levels remained significantly elevated (d = 0.16, p < 0.01) for a further 2 weeks, before returning to baseline levels. Coronavirus symptom severity explained changes in distress attributable to COVID-19, whereby distress was more pronounced among those whose symptoms were more severe and were slower to subside. Conclusions. This study indicates that testing positive for COVID-19 is associated with an initial increase in psychological distress that diminishes quickly as symptoms subside. Although COVID-19 may not produce lasting psychological distress among the majority of the general population it remains possible that a minority may suffer longer-term mental health consequences.