It is becoming more common for cochlear implant (CI) recipients to have some useful acoustic hearing post-operatively. Such CI users may benefit from simultaneous use of an acoustic hearing aid (HA). To provide a satisfactory fitting of both types of devices, it is helpful to make the loudness perceived via each device as similar as possible. In practice, this is usually attempted by adjusting the sensitivity and/or gain of each device during fitting, but it is often not clear whether or how well the loudness of the devices has been matched for each user. In this study, established numerical models of loudness for both acoustic and electric stimulation were used to investigate this problem. The models' outputs were compared to published psychophysical data and were found to approximate the main patterns observed in those data. They also showed that, theoretically, there would be differences in the loudness perceived with each type of device when sounds with varying frequency content and level were heard. Future sound processors and/or fitting techniques developed specifically for CI recipients with usable acoustic hearing could apply such computational models of loudness to improve simultaneous use of the two types of device.