BACKGROUND: Accurate diagnosis in patients with suspected coronavirus disease 2019 (COVID-19) is essential to guide treatment and limit spread of the virus. The combined nasal and throat swab is used widely, but its diagnostic performance is uncertain. METHODS: In a prospective, multi-centre, cohort study conducted in secondary and tertiary care hospitals in Scotland, we evaluated the combined nasal and throat swab with reverse transcriptase-polymerase chain reaction (RT-PCR) for severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in consecutive patients admitted to hospital with suspected COVID-19. Diagnostic performance of the index and serial tests was evaluated for a primary outcome of confirmed or probable COVID-19, and a secondary outcome of confirmed COVID-19 on serial testing. The diagnosis was adjudicated by a panel, who recorded clinical, laboratory and radiological features blinded to the test results. RESULTS: We enrolled 1368 consecutive patients (median age 68 [interquartile range, IQR 53-80] years, 47% women) who underwent a total of 3822 tests (median 2 [IQR 1-3] tests per patient). The primary outcome occurred in 36% (496/1368), of whom 65% (323/496) and 35% (173/496) had confirmed and probable COVID-19, respectively. The index test was positive in 255/496 (51%) patients with the primary outcome, giving a sensitivity and specificity of 51.4% (95% confidence interval [CI] 48.8 to 54.1%) and 99.5% (95% CI 99.0 to 99.8%). Sensitivity increased in those undergoing 2, 3 or 4 tests to 60.1% (95% CI 56.7 to 63.4%), 68.3% (95% CI 64.0 to 72.3%) and 77.6% (95% CI 72.7 to 81.9%), respectively. The sensitivity of the index test was 78.9% (95% CI 74.4 to 83.2%) for the secondary outcome of confirmed COVID-19 on serial testing. CONCLUSIONS: In patients admitted to hospital, a single combined nasal and throat swab with RT-PCR for SARS-CoV-2 has excellent specificity, but limited diagnostic sensitivity for COVID-19. Diagnostic performance is significantly improved by repeated testing.