Abstract: In order to detect the QTLs controlling the dynamic change of tobacco bacterial wilt (TBW) resistance, 158 F₆ recombinant inbred lines (RIL) derived from a cross between Dayemihe and Changbohuang were selected and genotyped with SSR and InDel markers, and a genetic map with 24 linkage groups was constructed by JoinMap 4 software. The map contained 546 SSR and 80 InDel markers and covered 2269.3 cM with a mean distance of 3.63 cM between adjacent markers. QTL mapping for TBW resistance was performed by WinQTLcart 2.5 software using the composite interval mapping (CIM) method based on TBW disease indexes at different survey time points in 2016 and 2017. 4, 4, 6 and 4 TBW resistance QTLs were detected at the four survey time points in 2016, respectively; the phenotypic variances of these QTLs ranged from 5.03% to 13.07%. At the four survey time points in 2017, 7, 3, 6 and 6 QTLs for TBW resistance were detected, respectively, and their phenotypic variances varied between 4.63% and 18.18%. A total of 28 QTLs were identified in two years, seven of them were detected repeatedly in different survey time points, but none could be found at all survey time points in both years. Moreover, the number and additive effect of TBW resistance QTLs had significant changes at four survey time points in both years. These results indicated that tobacco may utilize different TBW resistance genes at different stages during the pathogenesis process, and the expressions of these genes were time-dependent and sequential.