Communication commonly occurs with both linguistic and gestural signals. In spoken languages the gestural signal can be manual (e.g., meaningful hand gestures) or vocal (e.g., meaningful uses of pauses, volume, and intonation), but in signed languages non-linguistic gesture and language occupy the same visual–gestural channel. One type of gesture, constructed action, is characteristically mimetic and allows for depiction of a character's actions with the speaker's body. We examine that type of gesture as it complements narratives presented in American Sign Language (ASL) to various audiences. A single text (an account of a Deaf leader's life) was recounted by two native Deaf signers of ASL to three different audiences, a design that allows us to apply the sociolinguistic framework of style known as audience design (Bell, 1984). The data show that constructed action can occur in both non-formal and formal settings. Additionally, if constructed action is analyzed by body parts (i.e., head, torso, arms/hands, and legs/feet) and degree of production (i.e., slight, moderate, exaggerated), some trends appear across settings. We suggest these trends could be attributed to the signers accommodating to their audiences. Finally, we report an association between degree of emphasis of constructed action and audience/setting for the two signers.