Objectives To describe and compare the initial clinical characteristics of a cohort of patients with suspected COVID-19 managed by general practitioners (GPs); to assess whether 3-month persistent symptoms were more frequent among confirmed cases than among no-COVID cases; and to identify factors predictive of persistent symptoms and adverse outcomes among confirmed cases. Design and setting A comparative, prospective, multicentre cohort study in primary care in the Paris region of France. Participants 521 patients aged ≥18 with suspected COVID-19 were enrolled between March and May 2020. Outcome measures Initial symptoms, COVID-19 status, persistent symptoms 3 months after inclusion and a composite criterion for potentially COVID-19-related events (hospitalisation, death, emergency department visits). The final COVID-19 status (‘confirmed’, ‘no-COVID’ and ‘uncertain’ cases) was determined by the GP after the receipt of the laboratory test results. Results 516 patients were analysed; 166 (32.2%) were classified into the ‘confirmed COVID’ group, 180 (34.9%) into the ‘no-COVID’ group and 170 (32.9%) in the ‘uncertain COVID’ group. Confirmed cases were more likely to have persistent symptoms than no-COVID cases (p=0.09); initial fever/feeling feverish and anosmia were independently associated with persistent symptoms. At 3 months, we observed 16 (9.8%) COVID-19-related hospital admissions, 3 (1.8%) intensive care unit admissions, 13 (37.1%) referrals to an emergency department and no death. Age >70 and/or at least one comorbidity (OR 6.53; 95% CI 1.13–37.84; p=0.036), abnormalities in a lung examination (15.39; 95% CI 1.61–146.77; p=0.057) and two or more systemic symptoms (38.61; 95% CI 2.30–647.40; p=0.011) were associated with the composite criterion. Conclusions Although most patients with COVID-19 in primary care had mild disease with a benign course, almost one in six had persistent symptoms at 3 months. These symptoms were more frequent in the ‘confirmed COVID’ group. Our findings need to be confirmed in a prospective study with longer follow-up.