The name is known from Hecataeus to late antiquity and is found in dozens of authors (Hecat. fr. 18b; Hdt 4, 48; Thuc. 2, 96; Strab. 1, 3, 15; 7, 3, 13; 7, 3, 15; Plin. NH 4, 79; Iord. Get. 75).

The river is also mentioned by Hesiod, Pindar, Aeschylus, Hellanicus, Aristotle, Apollonius of Rhodes, Diodorus, Virgil, Ovid, and many other Greek and Roman historians and poets, as well as in inscriptions.

Tomaschek considers it probable that the name is derived from the root *sreu- “to flow”, accepting the initial [i] as “inorganic”. D. Detschew believes that Tomaschek incorrectly links Ἴστρος with the root *sreu- and accepts the etymology of Pokorni, who says that an initial form *Isros comes from the same root that we find in Ancient Greek ἱερός, Old Indian iširá-h “strong, mobile” and in the Old Celtic name Isгra. This base is *eis- “moving quickly, vigorously”, and the Greek adjective ἱερός “strong” Pokorni considers a variant of ἱαρός with a secondary e instead of a. I. Russu defines the name as Thraco-Getic and accepts the etymology from IE *eis-, as according to him the same root is also found in the hydronym Oiscos. V. Georgiev also derives the name from an IE root meaning “strong, turbulent,” but according to him it is *isro- (which is the same as *eis-, but with monophthongization) and defines this etymology as certain. Later, he writes that the original name of the Danube was *dehnu- *(h)ishro- “powerful (turbulent, large) river,” from which the Thracian “Ἴστρος,” the Celtic “Danuvius,” and the Daco-Mysian “Danavis” (which gave rise to the Romanian “Dunãre”) originated. This is a conjectural hypothesis, as there is no evidence that the name was ever composed of two parts. The Celtic and Thracian names of the river most likely arose completely independently of each other, and there is no evidence of a separate “Daco-Mysian” name.

Detschew draws the attention to the information provided by Stephanus of Byzantium about three other names of cities called ” Ἴστρος” besides the one at the mouth of the Danube – on the island of Crete, near the town of Knidos, and in Iapygia. Contemporary researchers believe that it was not near Knidos, but on a neighboring island, which was also called “Ἴστρος”, was located a city of the same name. As in other cases, we are dealing with an ancient Indo-European name, widespread in several languages on the Balkan Peninsula and in the Aegean basin. Since the list includes not only the river name but also names of islands and cities that are not located on rivers of the same name, I do not find acceptable, from a semantic point of view, the etymologies that derive it from roots meaning “to flow,” “to move quickly,” and “powerful, turbulent.” The probability of homonymy, in which the river name has one root and the settlement names have another, seems very small.