Abstract Background In the diagnosis of SARS-COV-2, upper respiratory tract specimens are primarily used for PCR testing. This study aims to evaluate the clinical and diagnostic role of bronchoalveolar lavage (BAL) in diagnosing and managing COVID-19. Methods Patients admitted to Keck Medical Center (Los Angeles, CA) during March 2020-May 2021 who had SARS-CoV-2 PCR tested on both nasopharyngeal swab (NPS) and BAL specimens (QIAGEN QIAstat-Dx RP) were included. Four patient groups were designated by PCR results: (A) NPS+/BAL+ (n=34), (B) NPS+/BAL- (n=26), (C) NPS-/BAL+ (n=5), (D) NPS-/BAL- (n=89). Clinical and laboratory data were retrospectively reviewed. Statistical analysis included the use of one-way ANOVA with Tukey HSD and Fisher’s exact test. Results 154 inpatients had NPS and BAL tested for SARS-CoV-2 RNA (mean age 55.1 y, 59.6% male). BAL was performed mainly to further assess abnormal respiratory function (n=43) or chest imaging findings (n=27), or for COVID-19 work-up (n=17). Mean turnaround time for BAL PCR was 4 h 5 min. Ct values, available on 28 BAL+ samples, did not significantly differ between patients who died (mean Ct 29.9, n=18) vs. survived (mean Ct 28.7, n=10). Time from symptom onset to BAL was significantly longer in Group B (34.2 d) vs. D (13.7 d, p< 0.01) or C (12.8 d, p< 0.05) but not vs. A (23.9 d). Groups did not differ significantly in patient comorbidities, mortality, length of hospital stay, or immunosuppressive status. SARS-CoV-2 NPS+/BAL+ patients were more likely to have diffuse or bilateral chest radiographic opacities (p< 0.01), respiratory culture with Aspergillus spp. (p< 0.05), and ARDS diagnosis than those who were NPS+/BAL- (p< 0.01). Conclusion BALs were an important specimen source for the detection of SARS-CoV-2 in the lower respiratory tract particularly when NPS at admission was PCR-negative and for ruling out COVID-19 pneumonia as a cause of abnormal pulmonary findings or declining respiratory status. BAL positivity for SARS-CoV-2 RNA was associated with higher frequency of diffuse or bilateral opacities on CXR, culture recovery of Aspergillus spp., and ARDS. Ct values of BAL+ samples did not predict mortality though sample size was limited. Further studies on the role of viral detection in BALs in patient management and prognosis are warranted. Disclosures All Authors: No reported disclosures.