Background: Molecular testing involving Epidermal growth factor receptor (EGFR) mutations is recommended for patients with non-small cell lung cancer (NSCLC), especially with lung adenocarcinoma, and is nowadays an essential element of precision medicine. Although amplification-refractory mutation system (ARMS) is one of the most commonly used methods for EGFR mutation detection, it can only detect variants in known gene loci and give indistinct results in gray area occasionally, which may be caused by the input of poor DNA quality or low-abundance variants. Idylla EGFR assay can detect 51 EGFR mutations directly from formalin-fixed, paraffin-embedded (FFPE) samples within 2.5 hours with < 2 minutes of hands-on time, without manual data analysis need, whose accurate and robust have been proven among Caucasian and Chinese patients. But the assay’s ability to detect EGFR mutation in low abundance which may lead the result of ARMS EGFR in gray area or rare variants has never been evaluated. Methods: FFPE samples were retrospectively collected from 98 lung cancer patients followed by EGFR testing using the Idylla system, which were also previously assessed with ARMS-PCR. Forty-six of them were in gray area by ARMS-PCR of EGFR and need to be retested by ARMS-PCR or validated by the third method. In another 52 cases, no mutations were found by ARMS-PCR of EGFR, but mutations were found by next-generation sequencing (NGS). The ability of Idylla EGFR to detect cases with result in gray area and rare variants were analyzed. Results: In 46 cases with result in gray area by ARMS-PCR of EGFR, after retesting or validation by the third method, 37 of them had the same result with Idylla mutation test, with an overall concordance of 80.4%. Idylla EGFR test missed 8 mutational cases (22/30), identified 1 mutational case in addition (15/16). Except 2 cases with higher CQ value than 29, which means the poor DNA quality or low quantity input, low mutational abundance maybe the main reason for no mutation identified by Idylla EGFR test. In 52 cases with rare EGFR variants, Idylla EGFR test identified 9 mutational cases, while 4/9 mutational variants beyond the scope of Idylla EGFR test, 2 20ins, 1 L858R and 1 19del. Five of the 9 mutational cases, Idylla test detected 5 different 19dels. The variants of all the 52 cases beyond the scope of ARMS-PCR of EGFR, and ARMS-PCR of EGFR didn’t find mutation. Conclusions: The Idylla system provides a rapid and accurate ability to differentiate result of ARMS-PCR of EGFR in gray area and identify more rare mutation variants which may decrease the repeat testing or validation. But the ability of NGS to find more actionable variants is still important.