This paper argues that there is a political, social, economic or even historical substratum to the location and development of settlement that is significant to the diachronic understanding of settlement systems but may easily get neglected in favour of discussions of settlement hierarchy or single events in settlement history. Hence, it postulates that the factors for initial site location may not be the same as the factors for further settlement development, and that said substratum should be explored to fully grasp the reasons behind settlement dynamics. In doing so, it focuses on two categories of sites - so-called bridge-sites at significant geographical locations and legionary garrisons turned colonies. Settlements used as examples are Horreum Margi, Naissus, Scupi, and Ratiaria in Moesia Superior.