Fundulus capensis Garman, 1895 was described from the unique holotype said to be from False Bay, Cape of Good Hope, South Africa. Largely ignored by killifish taxonomists, its classification has remained ambiguous for over a century. Radiography and computed tomography of the holotype reveal skeletal details that have been used in modern phylogenetic hypotheses of cyprinodontiform lineages. Osteological synapomorphies confirm it is a cyprinodontiform killifish and allow us to identify it to species. The first pleural rib on the second vertebra and a symmetrical caudal fin with hypural elements fused into a fan-shaped hypural plate corroborate its classification in the cyprinodontiform suborder Cyprinodontoidei. The twisted maxilla with an anterior hook and the premaxilla with an elongate ascending process both place it in the family Fundulidae. The pointed neurapophyses of the first vertebra that do not meet in the midline and do not form a spine exclude it from the family Poeciliidae. Presence of discrete exoccipital condyles excludes it from the subfamily Poeciliinae. Overall shape, position of fins, and meristic data agree well with those of the well-known North American killifish, F. heteroclitus. Fundulus capensis Garman, 1895, redescribed herein, is considered a subjective synonym of Fundulus heteroclitus (Linnaeus, 1766). Provenance of the specimen remains a mystery.