A biotic hypothesis for the eucalypt disease of unknown aetiolgy, Mundulla Yellows (MY), was tested by seeking non-host nucleic acids in affectedEucalyptus camaldulensistrees using high throughput sequencing (HTS). Ten contigs were selected from a pool of nucleic acids unique to diseased trees, and oligonucleotide probes were prepared for specific dot-blot hybridisation assay of these contigs. One of the contigs was found to be associated with a few MY-affected trees collected in South Australia and Western Australia, as well as with similarly affectedE. camaldulensistrees growing in Syria. The probes for this contig represented opposite polarities of the contig, and both hybridised with the same positive samples even after alkali treatment confirming that the contig that they represented was dsDNA. This contig, comprising 766 nt, showed 95% similarity to whole genome shotgun sequence ofLeifsonia xyli. This preliminary microbiome analysis of MY affected trees provides the first evidence that aLeifsoniaspecies may infect eucalypts and that it should be further studied as a potential agent of MY.