Novel large-effect consistent QTL for anther extrusion (AE) to improve cross-pollination were mapped in doubled haploid populations derived from IPK gene bank spring wheat accessions. TaAP2-D, an ortholog of Cleistogamy1 in barley, is a likely candidate gene for AE in wheat. To establish a robust hybrid wheat breeding system, male lines harboring alleles that promote outcrossing should be developed. In this study, we developed two doubled haploid (DH) populations of hexaploid spring wheat (Triticum aestivum L.) by crossing accessions taken from IPK gene bank. In both populations, the phenotypic data of anther extrusion (AE) based on three years of field trials showed a wide variation and approximated a normal distribution. Both populations were genotyped with a 15 k Infinium single nucleotide polymorphism (SNP) array resulting in 3567 and 3457 polymorphic SNP markers for DH population-1 and DH population-2, respectively. Composite interval mapping identified quantitative trait loci (QTL) on chromosomes 1D, 2D, 4A, 4B, 5A, 5D, 6A, and 6B; with consistent QTL (that are identified in all the years) on chromosome 4A in DH population-1, and on chromosomes 2D and 6B in DH population-2. The consistent QTL explained 17.2%, 32.9%, and 12.3% of the phenotypic variances, respectively. Genic scan of the chromosome 2D-QTL showed that the wheat gene TaAP2-D, an ortholog of Cleistogamy1 which promotes AE via swelling of the lodicules in barley, lies within the QTL region. A diagnostic marker was developed for TaAP2-D that showed co-segregation with the AE phenotype. This study shows the use of gene bank diversity reservoir to find alleles which are otherwise difficult to detect in elite populations. The identification of large-effect consistent QTL for AE is expected to help form efficient male parental lines suitable for hybrid wheat seed production and serve as a source for map-based cloning.