Although hybridization between non-sibling species rarely results in viable or fertile offspring, it occasionally produces self-perpetuating or sexually-parasitic lineages in which ancestral genomes are inherited clonally and thus may persist as ‘ghost species’ after ancestor extinction. Ghost species have been detected in animals and plants, for polyploid and diploid organisms, and across clonal, semi-clonal, and even sexual reproductive modes. Here we use a detailed investigation of the evolutionary and taxonomic status of a newly-discovered, putative ghost lineage (HX) in the fish genus Hypseleotris to provide perspectives on several important issues not previously explored by other studies on ghost species, but relevant to ongoing discussions about their detection, conservation, and artificial re-creation. Our comprehensive genetic (allozymes, mtDNA) and genomic (SNPs) datasets successfully identified a threatened sexual population of HX in one tiny portion of the extensive distribution displayed by two hemi-clonal HX-containing lineages. We also discuss what confidence should be placed on any assertion that an ancestral species is actually extinct, and how to assess whether any putative sexual ancestor represents a pure remnant, as shown here, or a naturally-occurring resurrection via the crossing of compatible clones or hemi-clones.