The adaptive significance of discontinuous gas exchange cycles (DGC) in insects is contentious. Based on observations of DGC occurrence in insects of typically large brain size and often socially-complex life history, and spontaneous DGC in decapitated insects, the neural hypothesis for the evolution of DGC was recently proposed. It posits that DGC is a non-adaptive consequence of adaptive down-regulation of brain activity at rest, reverting ventilatory control to pattern-generating circuits in the thoracic ganglia. In line with the predictions of this new hypothesis, we expected a higher likelihood of DGC in the gregarious phase of the desert locust (Schistocerca gregaria, Orthoptera), which is characterized by a larger brain size and increased sensory sensitivity compared with the solitary phase. Furthermore, surgical severing of the neural connections between head and thoracic ganglia was expected to increase DGC prevalence in both phases, and to eliminate phase-dependent variation in gas exchange patterns. Using flow-through respirometry, we measured metabolic rates and gas exchange patterns in locusts at 30°C. In contrast to the predictions of the neural hypothesis, we found no phase-dependent differences in DGC expression. Likewise, surgically severing the descending regulation of thoracic ventilatory control did not increase DGC prevalence in either phase. Moreover, connective-cut solitary locusts abandoned DGC altogether, and employed a typical continuous gas exchange pattern despite maintaining metabolic rate levels of controls. These results are not consistent with the predictions of the neural hypothesis for the evolution of DGC in insects, and instead suggest neural plasticity of ventilatory control.