Electrical cochlear implants (eCI) represent the most successful neuroprosthesis and allows for open speech comprehension in the majority of the more than 700 000 patients. Typically, in adult humans, eCIs are fitted to the individual patient via extensive patient feedback sessions in a multistage process. At the opposite, newborns eCI are fitted using a set of objective measures. Similar objective approach should be automated in animal models. The common marmoset (Callithrix jacchus) has become a prime model for auditory neuroscience in general and evaluation of eCI in particular. In the current study, marmosets were unilaterally deafened and implanted with an eCI. We evaluated two electrically evoked far-field potentials (electrically evoked auditory brainstem response, eABR; electrically evoked compound action potential, ECAP) as potential measures for marmoset eCI fitting and compared them with behavioral data from eCI stimulation. The eABR and ECAP correspond to the synchronous activation of the auditory pathway and the auditory nerve fibers, respectively. Both approaches are stable and widely used by clinicians to evaluate changes in the peripheral auditory system without requiring patient cooperation such as in children. Three marmosets implanted with eCI (deafened and implanted in the left ear) were tested with biphasic electrical pulses and artifact reduction methods. For the first time in this species, we obtained reliable, stable and reproducible ECAPs in two of our animals. Employing a Go-No Go procedure, behavioral thresholds for electrical intensity were estimated in two marmosets. These results were stable and reproducible for most channels. Moreover, preliminary analyses indicated that behavioral thresholds were reliably lower than ECAP thresholds, in accordance to human data. However, being time consuming, we think that this procedure could be accelerated by using ECAP thresholds. In conclusion, our work paves the way to future research on automatic fitting of eCI using ECAP thresholds.