The wheat dough quality is of great significance for the end-use of flour. Some genes have been cloned for controlling the protein fractions, grain protein content, starch synthase, grain hardness, etc. Using a unigene map of the recombinant inbred lines (RILs) for “TN 18 × LM 6,” we mapped a quantitative trait locus (QTL) for dough stability time (ST) and SDS-sedimentation values (SV) on chromosome 6A (QSt/Sv-6A-2851). The peak position of the QTL covered two candidate unigenes, and we speculated that TraesCS6A02G077000 (a xylanase inhibitor protein) was the primary candidate gene (named the TaXip gene). The target loci containing the three homologous genes TaXip-6A, TaXip-6B, and TaXip-6D were edited in the variety “Fielder” by clustered regularly interspaced short palindromic repeats–associated protein 9 (CRISPR/Cas9). Two mutant types in the T2:3 generation were obtained (aaBBDD and AAbbdd) with about 120 plants per type. The SVs of aaBBDD, AAbbdd, and WT were 31.77, 27.30, and 20.08 ml, respectively. The SVs of the aaBBDD and AAbbdd were all significantly higher than those of the wild type (WT), and the aaBBDD was significantly higher than the AAbbdd. The STs of aaBBDD, AAbbdd, and WT were 2.60, 2.24, and 2.25 min, respectively. The ST for the aaBBDD was significantly higher than that for WT and was not significantly different between WT and AAbbdd. The above results indicated that XIP in vivo can significantly affect wheat dough quality. The selection of TaXip gene should be a new strategy for developing high-quality varieties in wheat breeding programs.