Rapid development of high-throughput (HTS) molecular identification methods has revolutionized our knowledge about taxonomic diversity and ecology of fungi. However, PCR-based methods exhibit multiple technical shortcomings that may bias our understanding of the fungal kingdom. This study was initiated to quantify potential biases in fungal community ecology by comparing the relative performance of amplicon-free shotgun metagenomics and amplicons of nine primer pairs over seven nuclear ribosomal DNA (rDNA) regions often used in metabarcoding analyses. The internal transcribed spacer (ITS) barcodes ITS1 and ITS2 provided greater taxonomic and functional resolution and richness of operational taxonomic units (OTUs) at the 97% similarity threshold compared to barcodes located within the ribosomal small subunit (SSU) and large subunit (LSU) genes. All barcode-primer pair combinations provided consistent results in ranking taxonomic richness and recovering the importance of floristic variables in driving fungal community composition in soils of Papua New Guinea. The choice of forward primer explained up to 2.0% of the variation in OTU-level analysis of the ITS1 and ITS2 barcode data sets. Across the whole data set, barcode-primer pair combination explained 37.6–38.1% of the variation, which surpassed any environmental signal. Overall, the metagenomics data set recovered a similar taxonomic overview, but resulted in much lower fungal rDNA sequencing depth, inability to infer OTUs, and high uncertainty in identification. We recommend the use of ITS2 or the whole ITS region for metabarcoding and we advocate careful choice of primer pairs in consideration of the relative proportion of fungal DNA and expected dominant groups.