A fundamental property of human language is the distinction between lexical items that denote entities and those that describe events. While this noun-verb contrast is well-documented in spoken languages, less is known about whether and how it is marked in sign languages and whether iconicity and the interface to gesture play a special role in the grammatical distinction of word classes. Here, we investigate how German Sign Language (DGS) distinguishes nouns and verbs, focusing on the role of lexical demonstrations, that is, manual and non-manual gestural components that iconically depict meaning. We qualitatively examine morphophonological cues accompanying nominal and verbal uses of several potentially multifunctional DGS signs in data from a free elicitation task, as well as a sentence reproduction task. Qualitative analyses reveal systematic differences: Mouth gestures and constructed action (CA) predominantly accompany verbal uses, whereas mouthings (‘Mundbild’) and sequential size-and-shape specifiers (SASS) are characteristic of nominal uses of signs. Crucially, the choice of strategy is mediated by a sign’s iconic potential. Highly iconic signs are more likely to combine lexical content with gestural demonstrations in verbal contexts. Against this background, we argue that gestural demonstrations constitute a productive communicative resource that aligns with the noun-verb distinctions in DGS. Our analyses shed light on how grammar, iconicity, and gestural elements interact in the visual-spatial modality.