Somatic sensation is overwhelmingly felt as the result of mechanical stimulation or movement of the body or its parts. Indeed almost all tissues of the body receive an innervation from the peripheral axons of mechanosensitive sensory neurons with their cell body in the dorsal root or trigeminal ganglia. Remarkably, the mechanisms and molecules used by sensory neurons to transform mechanical force into an electrical signal are poorly understood. We term this process sensory neuron mechanotransduction, to distinguish it from mechanotransduction in the specialized epithelial hair cells of the inner ear. It is assumed that specialized mechanosensitive ion channels that are gated by force, underlie a graded receptor potential in sensory afferent endings. The graded receptor potential leads to action potential firing that decodes the stimulus magnitude. We have sought to be able to measure mechanosensitive currents in isolated sensory neurons in acute culture that presumably underpin the receptor potential. We have found two main types of highly sensitive mechanosensitive currents in the neurites and somas of sensory neurons (Lechner et al. 2009; Hu et al. 2010). These two currents have been termed rapidly-adapting and slowly adapting mechanosensitive currents. The rapidly adapting (RA-type current) is found in all mechanoreceptors but also in a substantial number of nociceptors and this current inactivates very rapidly. The slowly adapting current (SA-type) has a distinct pharmacology and biophysical profile from the RA-type current and is only found in nociceptive sensory neurons. The relevance of mechanosensitive currents for in vivo mechanosensitivity has been demonstrated by experiments showing that such currents are absent in neurons with a targeted disruption of the gene encoding the integral membrane protein called stomatin-like protein-3 (SLP3) (Wetzel et al, 2007). In SLP3 mutant mice a substantial number of sensory afferents in the skin completely lack mechanosensitivity. Recently we have addressed the mechanism by which the mechanosensitive current is activated in sensory neurons. In principle the current could be activated by membrane stretch or, analogous to the hair cell, an extracellular tether might transfer force from the matrix directly to the channel. We have used a variety of tools to manipulate extracellular proteins including limited proteolysis and combined physiological measurements with quantitative transmission electron microscopy. These experiments show that the presence of an extracellular tether protein filament with a length of 100 nm appears to be necessary for gating the RA-type current (Hu et al. 2010). The tether molecule is synthesized by sensory neurons and binds to a laminin-containing matrix. Our data is the first to show that a tether gating mechanism is relevant for somatic sensory neurons. Interestingly, activation of the SA-type current does not appear to depend on a link to the extracellular matrix. I will also present new data that the sensitivity of mechanosensitive channels in sensory neurons is maintained by unconventional motor proteins.