Highly　specific　sample　pretreatment　for　the　sensitive　detection　of　trace　bisphenol　A　(BPA)　in　compliacted　samples　is　critical.　Herein,　a　new　protocol　towards　online　specific　recognition　and　sensitive　detection　of　BPA　was　proposed　by　using　Aptamer@AuNPs-modified　affinity　monolith　coupled　with　LC-MS.　Optimization　of　polymerization　conditions　and　characterization　such　as　the　morphology,　energy　spectrum,　mechanical　stability,　aptamer　coverage　density　and　specific　performance　of　the　affinity　monolith　were　studied.　Nano-gold　particles　(AuNPs)　densely　distributed　on　the　rigid　hybrid-silica　substrates,　and　an　unusually　high　aptamer　coverage　density　reached　3388　pmol/mu　L,　which　was　favorable　to　fulfill　the　effective　identification　of　BPA　with　high　selectivity　and　inhibit　the　interference　of　analogs　including　BPB　and　BPC.　A　highly　sensitive　recognition　of　BPA　was　obtained　with　the　limit　of　detection　(LOD)　as　low　as　0.02　ng/mL.　Applied　to　dairy　milk　products　and　serum　samples,　trace　BPA　could　be　sensitive　detected　by　this　strategy,　while　the　poor　response　was　achieved　by　using　traditional　non-specific　SPE　column　for　sample　pretreatment.　Satisfactory　recoveries　of　fortified　BPA　were　measured　as　97.45　+/-　2.24%-98.03　+/-　4.36%　(n　=　3)　in　powdered　infant　formulas,　96.64　+/-　3.37%　similar　to　99.42　+/-　3.22%　(n　=　3)　in　bottled　milk,　94.69　+/-　2.15%　similar　to　100.96　+/-　1.94%　(n　=　3)　in　boxed　milk,　and　93.71　+/-　1.53%　similar　to　96.73　+/-　2.56%　(n　=　3)　in　children　serum　samples,　respectively.　This　protocol　lights　a　new　access　to　online　specific　identification　of　trace　BPA　from　complex　matrix　with　good　detection　sensitivity　by　using　aptamer-affinity　monolithic　column.