Compound nerve action potential recordings are often used intraoperatively to assist the surgeon in decision making on surgical approaches during nerve repair surgeries. The methods used clinically are based on surgical recordings (Robert et al., 2009). Although these methods are of great simplicity as handheld electrodes are applied to a nerve dissected and isolated from surrounding tissues, implementation of the methods in the clinical setting remains challenging. Here, we report findings from an experiment where similar recordings were made in nonhuman primate in order to improve our understanding and the performance of the methods. Recordings were made from the median nerve at the upper arm of an adult female pigtail monkey under TIVA (pentobarbital) and full NMB (pancuronium). Core body temperature was maintained near 38 °C. Nerve action potentials were induced by constant current stimulation and recorded using laboratory equipment. IONM electrodes were used: triple and double hook electrodes (Cadwell) for stimulation and recording, respectively. Using the same recording configuration as in the clinical situation, we failed to record recognizable compound action potentials across a short conduction length of a nerve (3–4 cm) that was isolated and suspended from surrounding tissues as large stimulus artifacts obscured the neuronal signal. The stimulus artifacts were suppressed and the compound nerve action potential became unmasked after a saline-soaked gauze was placed underneath the nerve between the stimulating and recording electrodes to bridge the nerve and the surrounding tissues (i.e., ‘bridge grounding’). Furthermore, compound action potentials were of small amplitude when all prongs of the stimulation electrode were placed underneath the nerve. However, the potentials increased markedly in size with another modification, i.e. when the nerve was placed between the middle prong and the outer 2 prongs of the stimulation electrode (i.e., ‘sandwich stimulation’) after reshaping of the prongs. The recordings were made at room temperature (about 23 °C) without use of warm saline irrigation, at which the compound action potentials preserved good amplitudes but slowed in conduction, leading to a better separation of compound action potentials from stimulus artifacts. We reversely translated the clinical methods of nerve compound action potential recording to a laboratory setting in order to better understand the factors affecting performance. We observed that compound nerve action potentials could be reliably recorded with IONM electrodes when 2 modifications, the ‘bridge grounding’ and ‘sandwich stimulation’, were made. Confirmation and additional experiments are underway to revisit the clinically used recording techniques in the laboratory setting.