Sunflowers (Helianthus annuus L.) are susceptible to multiple diseases in field production. In this study, we collected diseased sunflower leaves in fields located in South Dakota, USA, for virome investigation. The leaves showed visible symptoms on the foliage, indicating phomopsis and rust infections. To identify the viruses potentially associated with the disease diagnosed, symptomatic leaves were obtained from diseased plants. Total RNA was extracted corresponding to each disease diagnosed to generate libraries for paired-end high throughput sequencing. Short sequencing reads were assembled de novo and the contigs with similarities to viruses were identified by aligning against a custom protein database. We report the discovery of two novel mitoviruses, four novel partitiviruses, one novel victorivirus, and nine novel totiviruses based on similarities to RNA-dependent RNA polymerases and capsid proteins. Contigs similar to bean yellow mosaic virus and Sclerotinia sclerotiorum hypovirulence-associated DNA virus were also detected. To the best of our knowledge, this is the first report of direct metatranscriptomics discovery of viruses associated with fungal infections of sunflowers bypassing culturing. These newly discovered viruses represent a natural genetic resource from which we can further develop potential biopesticide to control sunflower diseases.