In this paper we present the multilingual dictionary of proper names Prolexbase, particularly the Serbian volume. We also present the complexity of proper names in the Serbian language, especially those related to their translation: their orthography, derivation, inflection and dialect variations. We describe the model of the Prolexbase, with the emphasize on the solutions that had to be made to include the Serbian language in the database (the use of two alphabets, several levels of derivations, the existence of multiple forms). At the end, we give some figures that corroborate the presence of the Serbian language in the Prolexbase.