The　effects　of　nuclear-data　uncertainties　on　the　predictions　of　the　reactor-physics　calculations　are　quantified　in　this　paper.　Uncertainty　analysis　has　been　applied　to　the　TMI-1　assembly　at　both　Hot　Zero　Power　(HZP)　and　Hot　Full　Power　(HET)　conditions,　propagating　the　nuclear-data　uncertainties　to　the　eigenvalue　and　few-group　constants.　The　statistical　sampling　method　has　been　applied　to　the　uncertainty　analysis　for　lattice　calculations　of　TMI-1　assembly.　The　172-group　WIMSD-4　formatted　microscopic　cross-section　library　and　corresponding　cross-section　covariance　matrices　based　on　the　ENDF/B-VII.1　have　been　generated　and　applied　for　the　uncertainty　analysis.　From　the　numerical　results　of　the　uncertainty　analysis,　it　can　be　observed　that　the　relative　uncertainties　of　the　eigenvalue　and　two-group　constants　at　HFP　are　a　little　larger　than　these　at　HZP.　The　v(t)　and　v(p)　covariance　matrices　of　U-235　impact　the　uncertainty　results　notably.　Both　at　HZP　and　HFP,　the　relative　uncertainty　of　the　eigenvalue　is　about　(4.7　+/-　0.09)parts　per　thousand　(applying　v(p)　covariance　of　U-235)　or　(7.1　+/-　0.20)parts　per　thousand　(applying　v(t)　covariance　of　U-235);　the　relative　uncertainties　of　two-group　constants　vary　from　about　(2.9　+/-　0.06)parts　per　thousand　(for　Sigma(t,2))　to　(11.9　+/-　0.34)parts　per　thousand　(for　Sigma(s,1,2)).　(C)　2017　The　Authors.　Published　by　Elsevier　Ltd.