ABSTRACT Gibberella zeae, causal agent of Fusarium head blight (FHB) of wheat and barley and Gibberella ear rot (GER) of corn, may be transported over long distances in the atmosphere. Epidemics of FHB and GER may be initiated by regional atmospheric sources of inoculum of G. zeae; however, little is known about the origin of inoculum for these epidemics. We tested the hypothesis that atmospheric populations of G. zeae are genetically diverse by determining the genetic structure of New York atmospheric populations (NYAPs) of G. zeae, and comparing them with populations of G. zeae collected from seven different states in the northern United States. Viable, airborne spores of G. zeae were collected in rotational (lacking any apparent within-field inoculum sources of G. zeae) wheat and corn fields in Aurora, NY in May through August over 3 years (2002 to 2004). We evaluated 23 amplified fragment length polymorphism (AFLP) loci in 780 isolates of G. zeae. Normalized genotypic diversity was high (ranging from 0.91 to 1.0) in NYAPs of G. zeae, and nearly all of the isolates in each of the populations represented unique AFLP haplotypes. Pairwise calculations of Nei's unbiased genetic identity were uniformly high (>0.99) for all of the possible NYAP comparisons. Although the NYAPs were genotypically diverse, they were genetically similar and potentially part of a large, interbreeding population of G. zeae in North America. Estimates of the fixation index (G(ST)) and the effective migration rate (Nm) for the NYAPs indicated significant genetic exchange among populations. Relatively low levels of linkage disequilibrium in the NYAPs suggest that outcrossing is common and that the populations are not a result of a recent bottleneck or invasion. When NYAPs were compared with those collected across the United States, the observed genetic identities between the populations ranged from 0.92 to 0.99. However, there was a significant negative correlation (R = -0.59, P < 0.001) between genetic identity and geographic distance, suggesting that some genetic isolation may occur on a continental scale. The contribution of long-distance transport of G. zeae to regional epidemics of FHB and GER remains unclear, but the diverse atmospheric populations of G. zeae suggest that inoculum may originate from multiple locations over large geographic distances. Practically, the long-distance transport of G. zeae suggests that management of inoculum sources on a local scale, unless performed over extensive production areas, will not be completely effective for the management of FHB and GER.