The　measurement　data　of　interval　harmonic　state　estimation　(IHSE)　in　power　system　is　composed　of　harmonic　monitoring　data　(HMD).　Therefore,　the　quality　of　HMD　is　important　to　the　accuracy　of　IHSE　in　power　system.　HMD　comes　from　the　power　quality　monitoring　device　(PQMD).　PQMD　detection　period　of　different　measuring　points　is　asynchronous.　This　paper　pointed　out　that　the　asynchronism　of　PQMD　may　cause　some　measurements　in　IHSE　to　deviate　from　the　actual　value　and　becomes　bad　data　when　the　system　state　changes　suddenly.　Large　amount　of　new　energy　is　connected　to　the　grid,　which　makes　the　state　of　the　grid　present　strong　randomness　and　volatility.　Therefore,　the　application　of　HMD　in　IHSE　is　restricted.　It　is　an　important　premise　and　basis　to　screen　the　bad　data　from　the　asynchronous　HMD　to　ensure　the　accuracy　of　IHSE.　According　to　the　characteristics　of　sampling　method　and　data　format　of　PQMD,　this　paper　pointed　out　that　the　bad　data　caused　by　the　asynchronism　of　PQMD　exists　in　the　mutational　data.　To　screen　out　the　bad　data,　a　two-step　screening　method　of　HMD　applied　to　IHSE　is　proposed　in　this　paper.　Firstly,　a　standardized　error　detection　method　based　on　the　self-variance　variation　rule　is　proposed.　This　method　can　screen　out　mutational　data　in　HMD.　Then,　a　data　screening　method　based　on　sliding　window　correlation　coefficient　is　proposed　to　screen　the　bad　data　in　the　mutational　data.　Finally,　the　effectiveness　of　the　proposed　method　is　verified　by　simulation　in　IEEE14-bus　system.　©　2023　IEEE.