Gibberella zeae, the principal cause of Fusarium head blight (FHB) of barley, contaminates grains with several mycotoxins, which creates a serious problem for the malting barley industry in the United States, China, and Europe. However, limited studies have been conducted on the trichothecene profiles and population genetic structure of G. zeae isolates collected from barley in the United States. Trichothecene biosynthesis gene (TRI)-based polymerase chain reaction (PCR) assays and 10 variable number tandem repeat (VNTR) markers were used to determine the genetic diversity and compare the trichothecene profiles of an older population (n = 115 isolates) of G. zeae collected in 1997 to 2000 with a newer population (n = 147 isolates) collected in 2008. Samples were from across the major barley-growing regions in North Dakota and Minnesota. The results of TRI-based PCR assays were further validated using a subset of 32 and 28 isolates of G. zeae by sequence analysis and gas chromatography, respectively. TRI-based PCR assays revealed that all the G. zeae isolates in both populations had markers for deoxynivalenol (DON), and the frequencies of isolates with a 3-acetyldeoxynivalenol (3-ADON) marker in the newer population were ≈11-fold higher than those among isolates in the older population. G. zeae populations from barley in the Midwest of the United States showed no spatial structure, and all the isolates were solidly in clade 7 of G. zeae, which is quite different from other barley-growing areas of world, where multiple species of G. zeae are commonly found in close proximity and display spatial structure. VNTR analysis showed high gene diversity (H = 0.82 to 0.83) and genotypic diversity but low linkage disequilibrium (LD = 0.02 to 0.07) in both populations. Low genetic differentiation (F(ST) = 0.013) and high gene flow (Nm = 36.84) was observed between the two populations and among subpopulations within the same population (Nm = 12.77 to 29.97), suggesting that temporal and spatial variations had little influence on population differentiation in the Upper Midwest. Similarly, low F(ST) (0.02) was observed between 3-ADON and 15-acetyldeoxynivalenol populations, indicating minor influence of the chemotype of G. zeae isolates on population subdivision, although there was a rapid increase in the frequencies of isolates with the 3-ADON marker in the Upper Midwest between the older collection made in 1997 to 2000 and the newer collection made in 2008. This study provides information to barley-breeding programs for their selection of isolates of G. zeae for evaluating barley genotypes for resistance to FHB and DON accumulation.