Abstract The present study demonstrates that the process of linguistic Romanization, i.e. Latinization of the Roman Empire, is traceable by the data of the Computerized Historical Linguistic Database of Latin Inscriptions of the Imperial Age (LLDB). A multi-level analysis of linguistic and non-linguistic data in the LLDB has shown that Latinization, i.e. the spread of spoken or vulgar Latin, became more and more intensive over time in all concerned provinces (i.e. Lusitania, Gallia Narbonensis, Venetia et Histria, Dalmatia, Moesia, Pannonia, and Britannia), although to a varying degree in each. What is more, in many aspects of the investigation, it was possible to find differences between the selected provinces of the Roman Empire corresponding mostly to the future Romance (both negative and positive) outcomes of the respective areas. All in all, the analysis of data of the LLDB database can contribute to solving the complex problem of Latinization, and is a lot more appropriate for this purpose than a simple comparative analysis of epigraphic corpora of the selected provinces.