Correct and fast identification of the agents of mycoses is of great importance to enable early diagnosis and targeted antifungal therapy. DNA barcoding using short DNA regions (barcodes) offers an accurate, fast, cost-effective, culture independent approach for species identification. The current primary fungal DNA barcode is the internal transcribed spacer (ITS) region. In 2015, the first quality controlled fungal barcode database for human and animal pathogenic fungi, the ISHAM-ITS database, was established. The database currently contains 4056 ITS sequences from 545 species. As the ITS1/2 region is only able to correctly identify 75% of all fungal species it prompted the need for a secondary DNA barcode locus to close the diagnostic capability gap. To fill this gap the translation elongation factor 1α (TEF1α) has recently been introduced as secondary fungal barcode and added to the ISHAM-ITS database. It currently contains 907 TEF1α sequences from 203 species. To assess the resolution power of TEF1α and ITS1/2 the distribution of inter-species to intra-species genetic distance within taxa sharing the same phylogenetic lineage has been compared. TEF1α showed less intra-species and higher discriminatory power at inter-species level than the ITS1/2, improving the identification success in taxa such as Scedosporium, Cryptococcus, or Meyerozyma. Recently, there has been a fundamental shift away from Sanger sequencing to Next Generation Sequencing (NGS) allowing for a high throughput sequence-based identification of complex samples (metabarcoding). One of those new technologies is the MinIONTM from Oxford Nanopore Technologies. It is a palm-sized sequencer with low initial start-up costs, a fairly simple sample preparation, is PCR free and fast. The MinIONTM offers much longer read lengths (average 10-15kb) and faster runs than amplicon-based sequencing, which is particularly important for rapid identification. To assess the advantages and pitfalls of nanopore sequencing in clinical diagnosis and to confirm the identification of Pneumocystis jirovecii directly from human sputum we used the MinIONTM. Total genomic DNA was extracted and sequenced using the Ligation Sequencing Kit 1D SQK-LSK108. After 6h of sequencing, 71305 reads were generated and analysed by Oxford Nanopore's cloud-based analysis platform, EPI2ME, where the WIMP workflow was used for taxonomic assignments. A total of 260 different taxa were identified from the 10997 classified reads. The majority of them represented bacteria. Only 25 fungal taxa were identified. This first application of the MinIONTM in fungal identification highlighted some major issues: (1) fungal taxa were represented by only a few reads (each 5-17), and (2) those showed only a low individual sequence identity (~90%), including P. jirovecii. Our results indicate that false-positive and error-prone reads currently represent a real challenge for metabarcoding studies. To overcome these issues, more accurate taxonomy assignment algorithms and reference databases are needed.