S eventeen years after the severe acute respiratory syndrome (SARS) epidemic, an outbreak of pneumonia, now called coronavirus disease (COVID-19), was reported in Wuhan, China. Some of the early casepatients had a history of visiting the Huanan Seafood Wholesale Market, where wildlife mammals are sold, suggesting a zoonotic origin. The causative agent was rapidly isolated from patients and identified to be a coronavirus, now designated as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by the International Committee on Taxonomy of Viruses (1). SARS-CoV-2 has spread rapidly to other places; 113,702 cases and 4,012 deaths had been reported in 110 countries/areas as of March 10, 2020 (2). In Hong Kong, 130 cases and 3 deaths had been reported. SARS-CoV-2 is a member of subgenus Sarbecovirus (previously lineage b) in the family Coronaviridae, genus Betacoronavirus, and is closely related to SARS-CoV, which caused the SARS epidemic during 2003, and to SARS-related-CoVs (SARSr-CoVs) in horseshoe bats discovered in Hong Kong and mainland China (3-5). Whereas SARS-CoV and Middle East respiratory syndrome coronavirus were rapidly traced to their immediate animal sources (civet and dromedaries, respectively), the origin of SARS-CoV-2 remains obscure. SARS-CoV-2 showed high genome sequence identities (87.6%-87.8%) to SARSr-Rp-BatCoV-ZXC21/ ZC45, detected in Rhinolophus pusillus bats from Zhoushan, China, during 2015 (6). A closer-related strain, SARSr-Ra-BatCoV-RaTG13 (96.1% genome identity with SARS-CoV-2), was recently reported in Rhinolophus affinis bats captured in Pu'er, China, during 2013 (7). Subsequently, Pangolin-SARSr-CoV/ P4L/Guangxi/2017 (85.3% genome identity to SARS-CoV-2) and related viruses were also detected in smuggled pangolins captured in Nanning, China, during 2017 (8) and Guangzhou, China, during 2019 (9). To elucidate the evolutionary origin and pathway of SARS-CoV-2, we performed an in-depth genomic, phylogenetic, and recombination analysis in relation to SARSr-CoVs from humans, civets, bats, and pangolins (10). The Study We downloaded 4 SARS-CoV-2, 16 human/civet-SARSr-CoV, 63 bat-SARSr-CoV and 2 pangolin-SARSr-CoV genomes from GenBank and GISAID (https:// www.gisaid.org). We also sequenced the complete genome of SARS-CoV-2 strain HK20 (GenBank accession no. MT186683) from a patient with COVID-19 in Hong Kong. We performed genome, phylogenetic, and recombination analysis as described (11). The 5 SARS-CoV-2 genomes had overall 99.8%-100% nt identities with each other. These genomes showed 96.1% genome identities with SARSr-Ra-BatCoV-RaTG13, 87.8% with SARSr-Rp-BatCoV-ZC45, 87.6% with SARSr-Rp-BatCoV-ZXC21, 85.3% with pangolin-SARSr-CoV/P4L/Guangxi/2017, and 73.8%-78.6% with other SARSr-CoVs, including human/civet-SARSr-CoVs (Table 1, https://wwwnc. cdc.gov/EID/article/26/7/20-0092-T1.htm). Most predicted proteins of SARS-CoV-2 showed high amino acid sequence identities with that of SARSr-Ra-BatCoV RaTG13, except the receptor-binding