The goal of this paper is to provide novel evidence revealing non-trivial constraints on the distribution of Slavic marked cardinals and indicating that such forms used as modifiers and predicates are in fact semantically complex expressions involving classifier semantics. I examine the evidence from Polish, Slovak and Bulgarian in comparison with classifier languages such as Mandarin Chinese and Japanese as well as languages with other languages with gendered numerals like Arabic. I propose a morphosemantic analysis grounded in Distributed Morphology and compositional semantics.