In this article, I present the Extended Accent First theory, which is an offshoot of the Primary Accent First theory (van der Hulst 1996, 1997, 1999, 2010). While the latter is known to correctly account for accent location in a large variety of languages, it encounters difficulties accounting for lexical accent systems and systems sensitive to both phonological weight and lexical accent. The Extended Accent First theory makes such an account possible. In this theory, lexical accent is reanalyzed as “diacritic weight”, leading to the notions of “diacritic weight scale” and “hybrid weight scale”. The Extended Accent First theory is illustrated here with a case study from Central and Southern Selkup that shows how the theory works and, in particular, how it can account for dominance effects using a diacritic weight scale. A comparison of the Accent Deletion approach vs. the Extended Accent First theory with respect to accentual dominance suggests that the approach proposed here is more straightforward and economical. Interestingly, the existence of phonological and diacritic weight correctly predicts that there are accent systems which make reference to both weight types (ordered in a single language-specific weight scale).