Infrared (wavelengths >750 nm) light-emitting equipment is commonly used worldwide to monitor nocturnal predator and prey behaviour. However, it is possible that the infrared (IR)-light wavelengths emitted from the equipment are so close to the spectral threshold of some key species that the light may be detected. An operant procedure was used to test whether five male ferrets (Mustela furo) could see an IR light with peak wavelengths of 870 and 920 nm. First, the ferrets were taught to press a lever under a lit white light for food reinforcement (overall mean response accuracy was 89%). Changing the properties (wavelength and intensity) of the light did not disrupt the ferrets’ abilities to perform the learned task. When the light was changed to IR (870 nm), four of five ferrets responded to the light at levels significantly higher than chance (mean = 68%, n = 4188, P < 0.01). When glare from a red trial-starting light was removed, two of the five ferrets (S3 and S4) showed strong evidence (response accuracies of 84% and 78%, respectively, P < 0.01) that they could see IR at 870 nm; however, S3 definitely could not see IR at 920 nm (n = 124, mean = 47%, P = 0.53). We conclude that at least some ferrets can see the light emitted from standard monitoring equipment that uses IR wavelengths of ~870 nm. To ensure nocturnal predator and prey behaviours are not altered by IR surveillance, field programs should use only high-wavelength IR diodes (at least 920 nm).