The adaptive nature of discontinuous gas exchange (DGE) in insects is contentious. The classic “hygric hypothesis”, which posits that DGE serves to reduce respiratory water loss (RWL), is still the best supported. We thus focused on the hygric hypothesis in this first ever experimental evolution study of any of the competing adaptive hypotheses. We compared populations of the migratory locust (Locusta migratoria) that underwent ten consecutive generations of selection for desiccation-resistance with control populations. Selected locusts survived 36% longer under desiccation stress but DGE prevalence did not differ between these and control populations (~75%). Evolved changes in DGE properties in the selected locusts included longer cycle and interburst durations. However, in contrast to predictions of the hygric hypothesis, these changes were not associated with reduced RWL rates. Other responses observed in the selected locusts were higher body water content when hydrated and lower evaporative water loss rates. Hence, our data suggest that DGE cycle properties in selected locusts are a consequence of an evolved increased ability to store water, and thus an improved capacity to buffer accumulated CO2, rather than an adaptive response to desiccation. We conclude that DGE is unlikely to be an evolutionary response to dehydration challenge in locusts.