Knowledge of Feline Leukemia Virus (FeLV) sequence variation has mainly been developed using Sanger sequencing methods. However, advances in next generation sequencing methods and their broad use in laboratories has been changing our understanding of viral genetics. FeLV sequencing has specific complications with the presence of both exogenous (exFeLV) and endogenous (enFeLV) virus with frequent recombination between them, limiting sequencing approaches. Here we report an FeLV-A and FeLV-B amplicon-based comparison between Sanger, Illumina, and Oxford Nanopore sequencing methods in Chilean domestic cats. We analysed the hypervariable envelope gene, where a higher number of variants as well as recombination with endogenous strains occurs. We detected multiple variants and viral quasispecies infecting the cats. We compared these three methods to evaluate the advantages and disadvantages between them. Although the Sanger method is highly reliable, it showed a high fail rate (many amplicons did not produce useable sequence) and the sequences obtained showed artificial sequence clustering when compared with the NGS methods. Illumina sequencing showed a lower error rate but could not discriminate between exogenous and endogenous viruses. Finally, Oxford Nanopore (MinION) sequencing could successfully detect low-abundance sequences and discriminate between FeLV-A and FeLV-B sequences, although its higher error-rate requires caution in interpretation of the results. Our results indicate advantages and disadvantages for each method, with the purpose of sequencing needing to be considered in the choice of method. Results of large viral phylogenetic trees combing sequences derived from mixed sequencing methods, such as those combining historical and contemporary sequencing need to be considered with some caution as artificial clustering by sequencing method may occur.