Genome editing poses inherent risks of genotoxicity that require thorough evaluation prior to clinical application. CAST-Seq, a genome-wide method for detecting CRISPR-Cas nuclease-induced chromosomal aberrations, has demonstrated high sensitivity but is less effective in identifying off-target activities of base editors (BEs), which are not designed to create DNA double-strand breaks (DSBs). To address this limitation, we developed Universal CAST-Seq (UCAST), a method that uses a DSB introduced by an auxiliary nuclease as an “anchor” to detect off-target effects associated with BEs. To prevent guide RNA cross-contamination, we selected a CCR5-targeting SaCas9 nuclease as the anchor nuclease. After validating UCAST using CRISPR-Cas9 nucleases with known off-target profiles, we applied it to assess the specificity of various BEs targeting the EMX1 locus in primary T cells. Amplicon sequencing of 39 off-target sites identified by UCAST revealed base editing at 3 off-target sites with TadCBEd, 10 with ABE8e, and 32 with evoCDA1-BE4max. Notably, only 1 of these sites showed off-target activity when cells were edited with the EMX1-targeting Cas9 nuclease. These results underscore that (i) off-target mutagenesis by BEs is challenging to predict, (ii) BEs can induce DSBs leading to chromosomal translocations, and (iii) Cas9 nucleases are unsuitable as surrogates for off-target detection of BEs. In conclusion, UCAST offers a robust, cost-effective, and time-efficient platform for genome-wide detection of off-target effects without requiring primer optimization. This makes UCAST especially valuable for early-stage safety screening of next-generation genome-editing tools.