Recent　years　have　witnessed　a　growing　investigation　of　terrestrial　laser　scanning　(TLS)　for　monitoring　the　deformation　of　tunnels.　TLS　provides　the　ability　to　obtain　a　more　accurate　and　complete　description　of　the　tunnel　surfaces,　allowing　the　determination　of　the　mechanism　and　magnitude　of　tunnel　deformation,　because　the　entire　surface　of　the　tunnel　is　more　concretely　modelled　rather　than　being　represented　by　a　number　of　points.　This　paper　models　and　analyses　the　point　clouds　from　TLS　to　detect　the　possible　deformation　of　a　newly　built　tunnel.　In　the　application　of　monitoring　the　Badaling　Tunnel　for　the　Winter　Olympics　2020　in　Beijing,　China,　the　proposed　method　includes　the　following　components:　the　tunnel　axis　is　automatically　estimated　based　on　a　3D　quadratic　form　estimation;　all　of　the　point　clouds　are　segmented　into　axis-based　blocks;　and　representative　points,　solved　by　a　singular　value　decomposition　(SVD)　method,　are　estimated　to　describe　the　tunnel　surface　and　establish　the　correspondence　of　data　between　days.　The　deformations　are　detected　in　the　form　of　the　distance　discrepancies　of　representative　points　and　verified　by　the　measurements　using　total　station.　©　Authors　2017.　CC　BY　4.0　License.