Abstract Reverse transcriptase-polymerase chain reaction (RT-PCR) testing is an important tool for diagnosing coronavirus disease 2019 (COVID-19). However, performance concerns have emerged recently, notably regarding sensitivity. We hypothesized that the clinical, biological, and radiological characteristics of patients with a false-negative first RT-PCR test and a final diagnosis of COVID-19 might differ from those of patients with a positive first RT-PCR test. We conducted a multicenter matched case–control study in COVID-19 patients. Patients with a negative first RT-PCR test were matched to patients with a positive first RT-PCR test on age, sex, and initial admission unit (ward or intensive care). We included 80 cases and 80 controls between March 30, and June 22, 2020. Neither mortality at hospital discharge nor hospital stay length differed between the two groups ( P = 0.80 and P = 0.54, respectively). By multivariate analysis, two factors were independently associated with a lower risk of a first false-negative test, namely, headache (adjusted OR [aOR], 0.07; 95% confidence interval [95% CI], 0.01–0.49]; P = 0.007) and fatigue/malaise (aOR, 0.16; 95% CI, 0.03–0.81; P = 0.027); two other factors were independently associated with a higher risk of a first false-negative test, namely, platelets > 207·10 3 mm −3 (aOR, 3.81; 95% CI, 1.10–13.16]; P = 0.034) and C-reactive protein > 79.8 mg·L −1 (aOR, 4.00; 95% CI, 1.21–13.19; P = 0.023). Patients with suspected COVID-19 whose laboratory tests indicating marked inflammation were at higher risk of a first false-negative RT-PCR test. Strategies involving serial RT-PCR testing must be rigorously evaluated.