Fusarium head blight (FHB) is an important disease of wheat (Triticum aestivum L.) that has caused billions of dollars in losses in recent decades. Although a massive breeding effort has been undertaken on multiple continents, there are no wheat cultivars with immunity to the disease. Resistance is conditioned by multiple loci and is further complicated by the role of the environment in expression of the disease phenotype. The objectives of our study were (i) to evaluate the phenotypic response to FHB in a large, diverse soft red winter wheat mapping panel; and (ii) to identify promising quantitative trait loci (QTL) associated with FHB resistance based on a genome‐wide association study (GWAS). We evaluated the mapping panel in 2014–2015 and 2015–2016 in an irrigated, inoculated scab nursery near Lexington, KY. Traits evaluated were heading date, plant height, FHB rating, severity, incidence, index, Fusarium‐damaged kernels (FDK), and deoxynivalenol (DON). There were significant (p < 0.05) differences among genotypes for all traits measured. The GWAS (based on 2‐yr entry means) identified 16 significant (p < 0.001) single nucleotide polymorphisms (SNPs) associated with disease traits on multiple chromosomes. Single nucleotide polymorphism association ranged from −2.14 to 4.01% of the mean of a given trait. We detected SNPs associated with FDK and DON on chromosomes 4A, 5B and 6B, and these SNPs decreased DON levels by 1.5, 2.1 and 3.2 mg kg−1, respectively. Our study demonstrated that even small‐effect QTL can potentially decrease disease levels and thus be useful in breeding programs.