This paper deals with ways in which gestural "pictures" are made, i.e., manual depictions of phenomena in the world. The view that "iconic" gestures uniformly function by way of some resemblance between signifier and signified is rejected, giving way to an understanding of depiction by gesture as the achievement of a heterogeneous set of practices, some of which rely on relations of contiguity or indexicality to evoke commonly known objects or scenes. Others seem to be derivative of other representation methods (e.g., drawing on surfaces). The paper reviews some existing work on gestural depiction methods, offers a working heuristics, and illustrates some of its categories. It is suggested that some of the basic ways in which actions of the hands evoke the world in gesture correspond to fundamental modes of existence and activity of human hands in the world: hands depict by enacting their familiar, "real-world" capacities as users, transporters, experiencers, assemblers, molders, and shapers of things.