This　paper　proposes　an　improved　metabolic　model　GM(1,　1)　in　order　to　enhance　forecasting　precision　for　life　of　steel　structures.　Based　on　the　complete　sequence,　using　the　new　data　taken　from　the　traditional　GM(1,　1)　model　to　replace　the　old　data,　an　improved　metabolic　GM(1,　1)　model　is　built,　which　not　only　ensures　the　original　dimensionality,　but　also　retains　the　growing　trend　of　the　whole　information.　Forecasting　for　coating　corrosion　in　a　concrete-filled　steel　tube　arch　bridge　with　the　proposed　model　is　carried　out　and　compared　with　the　traditional　GM(1,　1)　model.　It　is　found　that　the　mean　square　error　C　with　the　improved　model　is　0.132　9,　which　is　much　lower　than　the　value　0.172　1　obtained　with　the　traditional　GM(1,　1)　model.　The　relative　mean　error　of　the　improved　metabolic　GM(1,　1)　model(3.20%)　is　also　smaller　than　that　of　the　traditional　metabolic　GM(1,　1)　model(4.01%).　Conclusion　is　that　the　improved　metabolic　GM(1,　1)　model　has　higher　forecasting　accuracy　than　the　traditional　one.　It　ensures　the　original　dimensionality　and　retains　the　growing　trend　of　the　whole　information,　which　means　the　improved　metabolic　model　GM(1,　1)　is　more　reasonable　and　suitable　for　long　term　forecasting.　©　2015,　Editorial　Department　of　Journal　of　SJZU.　All　right　reserved.