Recent studies from multiple labs including ours have demonstrated the importance of extrachromosomal circular DNA (eccDNA) from yeast to humans (Shibata et al., 2012; Dillon et al., 2015; Møller et al., 2016; Kumar et al., 2017; Turner et al., 2017; Kim et al., 2020). More recently, it has been found that cancer cells obtain a selective advantage by amplifying oncogenes on eccDNA, which drives genomic instability (Wu et al., 2019; Kim et al., 2020). Previously, we have purified circular DNA and enriched the population using rolling circle amplification followed by high-throughput sequencing for the identification of eccDNA based on the unique junctional sequence. However, eccDNA identification by rolling circle amplification is biased toward small circles. Here, we report a rolling circle-independent method to detect eccDNA in human cancer cells. We demonstrate a sensitive and robust step-by-step workflow for finding novel eccDNAs using ATAC-seq (Assay for Transposase-Accessible Chromatin using sequencing) combined with a Circle_finder bioinformatics algorithm to predict the eccDNAs, followed by its validation using two independent methods, inverse PCR and metaphase FISH (Fluorescence in situ Hybridization).