Rust diseases are a significant threat to wheat crops, and they can easily reach epidemic proportions due to their very rapid asexual infection cycle, which produces prodigious numbers of dikaryotic urediniospores. Each urediniospore contains two nuclei, each with one haploid genome. The genetic organization of urediniospores leads to evolutionary success through frequent virulence shifts, often defeating resistance genes deployed in commercial wheat cultivars via both mutations and (a)sexual recombination. Furthermore, when new predominant races appear, they quickly spread through rapid asexual reproduction throughout growing seasons. These dikaryotic species bring challenges to analyzing the structural and repetitive regions of rust genome architecture with short-read sequencing. To resolve the challenges, long-read sequencing that requires pure high molecular weight (HMW) DNA is now applied to generate more contiguous rust genomes with haplotype-specific information. However, extraction of HMW DNA from wheat rust spores is especially difficult due to restricted biomass and high content of polysaccharides. Here, an improved DNA extraction protocol for rust spores/urediniospores is described that provides a substantial yield of pure HMW DNA. The average molecular weight of isolated DNA is larger than 50 kilobase pairs (kbp) and is sufficient for long-read sequencing technologies, such as Pacific Biosciences and Nanopore that enable increases in genome resolution and resolution of complex loci and nucleotide polymorphisms to explore rust genetic variation and evolution.