The number of cochlear implant (CI) users is increasing annually, resulting in an increase in the workload of implant centers in ongoing patient management and evaluation. Remote testing of speech recognition could be time-saving for both the implant centers as well as the patient. This study addresses two methodological challenges we encountered in the development of a remote speech recognition tool for adult CI users. First, we examined whether speech recognition in noise performance differed when the steady-state masking noise was presented throughout the test (i.e. continuous) instead of the standard clinical use for evaluation where the masking noise stops after each stimulus (i.e. discontinuous). A direct coupling between the audio port of a tablet computer to the accessory input of the sound processor with a personal audio cable was used. The setup was calibrated to facilitate presentation of stimuli at a predefined sound level. Finally, differences in frequency response between the audio cable and microphones were investigated.