Modern studies in Roman law pay much attention to Roman predial servitudes and their origin traced back to the Twelve Tables. However, traditionally the origin of servitudes is connected with Roman private law and, consequently, with private ownership of land, though this does not agree very well with the fact that in the archaic period the land was almost exclusively in public property. At the same time, one can infer from the sources that the Romans had public predial servitudes (such as via and aquae haustus), though some modern scholars maintain that they did not exist till the post-classical law and were unknown to the classical jurists. No one denies, however, that many personal servitudes, namely, superficies and emphyteusis, developed not later than in the republican period as institutes of public law and were transferred to the sphere of private law in the imperial period. But again, no one is ready to admit that servitutes praediorum could undergo similar evolution. Meanwhile, the sources make it possible to conclude that classical Roman jurists and agrimensors used the term praedium to denote both private and public land and employed the term of predial servitude when referring to public law relations. Moreover, the data of Liber coloniarum, those of municipal laws of the 1st c. BC, and the fragments of Hyginus Gromaticus allow the author to assert that the public servitude of via existed as early as in the republican time. In addition, analysing the information referring to the early Roman international agreements and to the history of Via Appia, which in the 4th c. BC ran from Rome to Capua through the lands belonging not only to the Romans but also to their allies, the author comes to the conclusion that servitude of via was developing in the sphere of international law. An analysis of the Digest and Livy allows the author to assume that the servitus cloacae immittendae was originally public in character, and the agrarian law of 111 BC describes servitus pecoris pascendi as a public servitude. The final conclusion of the author is that some kinds of servitudes came about earlier than private property of land and were developing originally as public ones; they were transferred to the sphere of private law only when private ownership of land had become widespread.