Although the neurophysiological basis of insect flight control has been studied extensively and successfully in animals attached to rigid tethers, these conditions disrupt the natural feedback between the subject's intentions, sensory input, and motor output. Understanding how individual control algorithms are integrated at a behavioral level requires acquisition and modification of biopotentials in completely untethered, free-flying animals. Herein, I present and test a miniaturized animal-computer interface for use with freely-flying Manduca sexta hawkmoths. This device is capable of simultaneously acquiring two independent biopotential signals, applying electrical neuromuscular stimulation, and correlating collected and applied signals with behavioral data from high-speed videography. Application of this device may offer substantial insight into how insects fly and, by replicating these mechanisms, facilitate wider application of micro air vehicles through improved flight efficiency, stability, and maneuverability.