The Abyssinian pea (Pisum abyssinicum A. Br.), concerned in this review, is known from Ethiopia and Yemen, where it is cultivated along with the common pea (Pisum sativum L. subsp. sativum). The continuously reproduced notion of its possible spontaneous occurrence in the wild ascends to suppositions made in the XIX century and is not based on any actual data. P. abyssinicum is of practical interest owing to its extra early ripening and resistance to bacterial blight. Morphologically it is very similar to P. sativum but its crossability with it is bad as either seed or pollen parent. Traditionally this reproductive barrier was associated with karyological differences. The Abyssinian pea karyotype is variable as 1-2 reciprocal translocations were reported. At the same time there are accessions not differing from the standard karyotype of P. sativum with respect to reciprocal translocations, yet their crossability with the latter is very low and the pollen fertility of F1 and F2 hybrids is lowered. Data were reported on influence of the region of Linkage Group III, containing a gene known to participate in the conflict of nucleus and plastids in remote crosses of peas, on the pollen fertility of hybrids with abyssinian pea. With their karyological variability, the known accessions of the Abyssinian pea are very close to each other genetically, as they diverged just about 4 000 years ago. The presence of alleles of molecular markers common with Pisum fulvum Sibth. et Smith on the one hand and P. sativum L. subsp. elatius (Bieb.) Schmalh. on the other hand evidences in favour of an old hypotheses by L.I. Govorov that the Abyssinian pea originated from their spontaneous hybrid. This spontaneous cross may have taken place under cultivation, in Yemen or Afar Depression. A representative of P. sativum subsp. elatius was revealed, the F1 hybrids of which with the Abyssinian pea as a seed parent had fully fertile pollen. P. abyssinicum. P. fulvum crosses provide the best hybrid seed outcome among remote crosses conducted, so that P. abyssinicum can be used as a 'bridge' for gene introgression from P. fulvum to P. sativum. Rather a high level of reproductive isolation of the Abyssinian pea from other representatives of the genus conforms the biological species concept, however the disposition of P. abyssinicum accessions as a small cluster among accessions of P. sativum subsp. elatius on molecular phylogeny reconstructions violates the phylogenetic species concept. Most authors assume the Abyssinian pea as a species, Pisum abyssinicum, some as a subspecies, P. sativum subsp. abyssinicum (A. Br.) Berger. Perhaps it would be most correct to consider it as a hybridogenic species. Because of the recent subsuming of the genus Pisum L. into the genus Lathyrus and with respect to the existing name Lathyrus abyssinicus A. Br. (a synonym of L. sativus L.), the Abyssinian pea is given a new name Lathyrus schaeferi (A. Braun) Kosterin nomen novum pro Pisum abyssinicum A. Braun), in honour of Hanno Schaefer, who substantiated the revision of tribe Fabeae by molecular reconstruction of its phylogeny. New combinations of Lathyrus sectio Pisum (L.) Kosterin combinatio nova and Lathurus fulvus (Sibthrop et Smith) Kosterin combinatio nova are proposed.