While normal hearing listeners are able to localize sound sources in the azimuthal plane accurately, bilateral cochlear implant (CI) users perform significantly worse in this task. Multiple reasons have been identified and are both processor and brain related. However, some aspects are still not completely understood, especially the complex interplay between the different stages. While the long-term goal is to shed more light on the overall process, the present study starts by investigating the salience of binaural cues encoded in the electrodogram. This was done by extracting the interaural level difference (ILD) and interaural time difference (ITD) of HRTF filtered speech directly after a simulation of CI encoding. As expected, the ILD can be well encoded in the absence of automatic gain controls. For the preservation of the ITD, a low-pass filtering is required, even after a Hilbert envelope extraction. Furthermore, an interaural electrode mismatch degrades both ILD and ITD but in different ways. It will be discussed if spread of excitation can help to reduce these mismatches. In a next step the perceptual relevance of these degradations is going to be investigated using an auditory processing model and/or psychoacoustic experiments.