SARS-CoV-2 RT-PCR with pooled specimens has been implemented during the COVID-19 pandemic as a cost- and manpower-saving strategy for large-scale testing. However, there is a paucity of data on the efficiency of different nucleic acid extraction platforms on pooled specimens. This study compared a novel automated high-throughput liquid-based RNA extraction (LRE) platform (PHASIFYTM) with a widely used magnetic bead-based total nucleic acid extraction (MBTE) platform (NucliSENS® easyMAG®). A total of 60 pools of nasopharyngeal swab and 60 pools of posterior oropharyngeal saliva specimens, each consisting of 1 SARS-CoV-2 positive and 9 SARS-CoV-2 negative specimens, were included for the comparison. Real-time RT-PCR targeting the SARS-CoV-2 RdRp/Hel gene was performed, and GAPDH RT-PCR was used to detect RT-PCR inhibitors. No significant differences were observed in the Ct values and overall RT-PCR positive rates between LRE and MBTE platforms (92.5% (111/120] vs 90% (108/120]), but there was a slightly higher positive rate for LRE (88.3% (53/60]) than MBTE (81.7% (49/60]) among pooled saliva. The automated LRE method is comparable to a standard MBTE method for the detection of SAR-CoV-2 in pooled specimens, providing a suitable alternative automated extraction platform. Furthermore, LRE may be better suited for pooled saliva specimens due to more efficient removal of RT-PCR inhibitors.