SUMMARY Polyphenic development is thought to play a pivotal role in the origin of morphological novelties. However, little is known about how polyphenisms evolve in natural populations, the developmental mechanisms that may mediate such evolution, and the consequences of such modification for patterns of morphological variation. Here we examine the developmental mechanisms of polyphenism evolution in highly divergent natural populations of the dung beetle, Onthophagus taurus. Males of this species express two alternative morphologies in response to larval feeding conditions. Favorable conditions cause males to grow larger than a threshold body size and to develop a pair of horns on their heads. Males that encounter relatively poor conditions during larval life do not reach this threshold size and remain hornless. Exotic populations of O. taurus have diverged dramatically in body size thresholds in less than 40 years since introduction to new habitats, resulting in the expression of highly divergent and novel horn length–body size scaling relationships in these populations. Here we show that larvae of populations that have evolved a larger threshold body size (1) have to accumulate greater mass to become competent to express the horned morph, (2) require more time to complete the final instar, (3) are less sensitive to the juvenile hormone (JH) analogue methoprene, and (4) exhibit a delay in the sensitive period for methoprene relative to other developmental events. JH has been shown previously to control horn expression in this species. Our results show that threshold evolution may be mediated via changes in the degree and timing of sensitivity to JH and may result in correlated changes in the dynamics and duration of larval development. Strain-specific differences in JH sensitivity have previously been demonstrated in other insects. However, to the best of our knowledge this is the first demonstration that changes in the timing of the sensitive period for JH may play an equally important role in the evolution of novel thresholds. We discuss our findings in the context of the developmental regulatory mechanisms that underlie polyphenic development and use our results to explore the consequences of, and constraints on, polyphenism evolution in nature.