Genetic approaches in several organisms provide the means of solving a previously intractable problem: characterizing the molecular foundations of the mechanical senses. In nematode mechanosensory cells, members of a novel class of epithelial ion channel subunits have been implicated as components of a mechanically gated channel. In insect mechanosensory bristles, mutations specifically defective in mechanoreceptor potentials have been identified. And in bacteria, a stretch-activated channel has been molecularly characterized for the first time. Although mechanosensitivity can be a property of an isolated channel, sensory transduction in eukaryotic mechanosensory cells probably requires the interaction of several membrane and cytoskeletal components.