The mobility hypothesis could explain the evolution of female‐biased size dimorphism if males with a smaller body size and longer legs have an advantage in scramble competition for mates. This hypothesis is tested by performing a selection analysis in the wild on Micrarchus hystriculeus (Westwood) (Phasmatodea), a sexually size dimorphic stick insect endemic to New Zealand. This analysis examined the form and strength of sexual selection on body size, leg lengths (front, mid and hind), and clasper size (a genitalic trait), and also quantified the degree of phenotypic variation and the allometric scaling pattern of these traits. By contrast to the mobility hypothesis, three lines of evidence were found to support significant stabilizing sexual selection on male hind leg length: a significant nonlinear selection gradient, negative static allometry, and a low degree of phenotypic variation. Hind leg length might be under stabilizing selection in males if having average‐sized legs facilitates female mounting or improves a male's ability to achieve the appropriate copulation position. As predicted, a negative allometric scaling pattern and low phenotypic variation of clasper size is suggestive of stabilizing selection and supports the ‘one‐size‐fits‐all’ hypothesis. Opposite to males, the mid and hind leg lengths of females showed positive static allometry. Relatively longer mid and hind leg lengths in larger females might benefit individuals via the better support of their larger abdomens. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 471–484.