Sensory neurons embedded in skin are responsible for the sense of touch. In humans and other mammals, touch sensation depends on thousands of diverse somatosensory neurons. By contrast, Caenorhabditis elegans nematodes have six gentle touch receptor neurons linked to simple behaviors. The classical touch assay uses an eyebrow hair to stimulate freely moving C. elegans, evoking evasive behavioral responses.While this assay has led to the discovery of genes required for touch sensation, it does not provide control over stimulus strength or position. Here, we present an integrated system for performing automated, quantitative touch assays that circumvents these limitations and incorporates automated measurements of behavioral responses. Highly Automated Worm Kicker (HAWK) unites microfabricated silicon force sensors and video analysis with real-time force and position control. Using this system, we stimulated animals along the anterior-posterior axis and compared responses in wild-type and spc-1(dn) transgenic animals, which have a touch defect due to expression of a dominant-negative a spectrin protein fragment.As expected from prior studies, delivering large stimuli anterior to the mid-point of the body evoked a reversal, but such a stimulus applied posterior to the mid-point evoked a speed-up. The probability of evoking a response of either kind depended on stimulus strength and location; once initiated, the magnitude and quality of both reversal and speed-up behavioral responses were uncorrelated with stimulus location, strength, or the absence or presence of the spc-1(dn) transgene. Wild-type animals failed to respond when the stimulus was applied near the mid-point. These results establish that stimulus strength and location govern the activation of a stereotyped motor program and that the C. elegans body surface consists of two receptive fields separated by a gap.