The paper will provide an overview of the various domains covered by TEI Lex- 0 and the main decisions that were taken over the last 18 months: constraining the general structure of a lexical entry; offering mechanisms to overcome the limits of <entry> when used in retro-digitized dictionaries (by allowing, for instance, <pc> and <lbl> as children of <entry>); systematizing the representation of morpho-syntactic information; providing a strict <sense>-based encoding of sense-related information; deprecating <hom>; dealing with internal and external references in dictionary entries, providing more advanced encodings of etymology; as well as defining technical constraints on the systematic use of @xml:id at different levels of the dictionary microstructure.