Cooper cable is made for implanted devices, usually for connection to stimulating electrodes. An experiment has been performed to see whether these cables would be satisfactory for recording electroneurogram (ENG) signals from cuffs. Four cables were subjected to continuous flexion at 2 Hz while submerged in saline. The cables were connected to a low-noise amplifier, and the noise was measured using a spectrum analyser. These cables had not fractured after 184 million flexions, and the noise in the neural band (500-5000 Hz) had not increased owing to age. Noise in the ENG band increased by less than 3 dB owing to the motion. A fifth, worn cable did fail during the experiment, the conductors becoming exposed to the saline, but this was only apparent by extra noise when the cable was in motion. After 184 million flexions, the four cables were given a more severe test: instead of being connected to the amplifier reference node, two of the four cores of each cable were connected to 18V batteries. Two of the cables were then noisier, but only when in motion, presumably because of leakage between cores. Cooper cables are excellent for transmitting neural signals alone; transmission in one cable of neural signals and power supplies should be avoided if possible.