High-throughput sequencing (HTS) identifies random viral fragments in environmental samples metagenomically. High reliability gains it broad application in virus evolution, host-virus interaction, and pathogenicity studies. Deep sequencing of field samples with content of host genetic material and bacteria often produces insufficient data for metagenomics and must be preceded by target enrichment. The main goal of the study was the evaluation of HTS for complete genome sequencing of field-case rabies viruses (RABVs). The material was 23 RABVs isolated mainly from red foxes and one European bat lyssavirus-1 isolate propagated in neuroblastoma cells. Three methods of RNA isolation were tested for the direct metagenomics and RABV-enriched approaches. Deep sequencing was performed with a MiSeq sequencer (Illumina) and reagent v3 kit. Bioinformatics data were evaluated by Kraken and Centrifuge software and de novo assembly was done with metaSPAdes. Testing RNA extraction procedures revealed the deep sequencing scope superiority of the combined TRIzol/column method. This HTS methodology made it possible to obtain complete genomes of all the RABV isolates collected in the field. Significantly greater rates of RABV genome coverages (over 5,900) were obtained with RABV enrichment. Direct metagenomic studies sequenced the full length of 6 out of 16 RABV isolates with a medium coverage between 1 and 71. Direct metagenomics gives the most realistic illustration of the field sample microbiome, but with low coverage. For deep characterisation of viruses, e.g. for spatial and temporal phylogeography during outbreaks, target enrichment is recommended as it covers sequences much more completely.