Recent advances in the field of environmental DNA (eDNA) metabarcoding have produced a promising platform for early detection and broad‐spectrum monitoring of plant pests and pathogens. To date, the majority of fungal metabarcoding assays have relied upon universal primers amplifying segments of the ribosomal DNA, to ensure broad taxonomic coverage while studying changes in community composition, or more recently, to screen for presence/absence of target species. In a diagnostics framework however, single, universal molecular barcodes do not discriminate accurately enough for many groups of important fungal phytopathogens at the required species level. Here, a modular, multi‐barcode amplicon sequencing pipeline was established to provide rapid and reliable diagnostics targeting the Ophiostomatales, an order containing economically important phytopathogens vectored by bark and ambrosia beetles (Curculionidae). Using compositionally varied mock communities, we evaluated five barcoding loci: ITS1, ITS2, the large ribosomal sub‐unit (LSU), translation elongation factor 1‐alpha (TEF1α) and calmodulin (CAL), for their ability to provide species‐level resolution. The sensitivity and detection limit of the assay was established by spiking genomic DNA of the exotic Dutch elm disease pathogen, Ophiostoma novo‐ulmi Brasier, into both the controlled mock communities and eDNA sampled from Ips grandicollis (Eichhoff) beetles collected in New South Wales, Australia. The TEF1α barcode successfully detected the Dutch elm pathogen DNA with 100% accuracy down to approx. 5 pg/μl, while the ITS1 barcode had a 100% accuracy down to approx. 0.4 pg/μl within a given sample. Our results demonstrate that a dual‐barcode approach targeting the ITS1 and TEF1α regions simultaneously is sufficient for accurate species level detection and characterisation of the Ophiostomatales and offers confident detection of low‐abundance taxa (relative read abundances of 0.1%–0.01%). Multi‐barcode metabarcoding provides a framework which can quickly and efficiently be adapted to new targets when establishing high throughput diagnostics for post‐border biosecurity surveillance of fungal phytopathogens.