Accurate　estimation　of　aircraft　fuel　consumption　is　critical　for　airlines　in　terms　of　safety　and　profitability.　In　current　practice,　estimation　of　fuel　consumption　for　a　flight　trip　is　usually　done　by　engineering　approaches,　which　mainly　consider　physical　factors,　e.g.,　planned　weather　and　planned　cruise　level.　However,　the　actual　performance　of　a　flight　usually　deviates　from　such　estimation.　Therefore,　we　propose　a　novel　self-organizing　constructive　neural　network　(CNN)　that　features　a　cascade　architecture　and　analytically　determines　connection　weights　to　estimate　the　trip　fuel　of　a　flight.　The　proposed　method　generates　non-redundant　and　linearly　independent　hidden　units　by　an　orthogonal　linear　transformation　of　operational　parameters　to　achieve　the　best　least-squares　solution.　Our　findings　provide　insights　for　the　aviation　industry　in　controlling　airlines＇　excess　fuel　consumption.