Vocal production learning is the ability to modify a vocal output in response to auditory experience. It is essential for human speech production and language acquisition. Vocal learning evolved independently several times in vertebrates, indicating evolutionary pressure in favor of this trait. This enables cross-species comparative analysis to be used to test evolutionary hypotheses. Humans share this ability with a versatile but limited group of species: songbirds, parrots and hummingbirds, bats, cetaceans, seals, and elephants. Although case studies demonstrate that African savanna and Asian elephants are capable of heterospecific imitation, including imitation of human words, our understanding of both the underlying mechanisms and the adaptive relevance within the elephant's natural communication system is limited.Even though comparing phylogenetically distant species is intriguing, it is also worthwhile to investigate whether and to what extent learned vocal behavior is apparent in species phylogenetically close to an established vocal learner. For elephants, this entails determining whether their living relatives share their special ability for (complex) vocal learning.In this review, we address vocal learning in Elephantidea and Sirenia, sister groups within the Paenungulata. So far, no research has been done on vocal learning in Sirenians. Because of their aquatic lifestyle, vocalization structure, and evolutionary relationship to elephants, we believe Sirenians are a particularly interesting group to study. This review covers the most important acoustic aspects related to vocal learning in elephants, manatees, and dugongs, as well as knowledge gaps that must be filled for one to fully comprehend why vocal learning evolved (or did not) in these distinctive but phylogenetically related taxa.