Music and prosody share many of the same relevant acoustic features: pitch, rhythm, and timbre. For cochlear implant (CI) users perception of all of these is challenging. While previous studies have primarily examined implant outcome in adult CI-recipients with an acquired hearing loss (HL), few studies have investigated perception of music, prosody, and speech in the growing population of adolescent CI users with a congenital HL. However, recent studies indicate that to keep pace with their normal hearing (NH) peers, supplementary measures of rehabilitation are important throughout adolescence. This study aimed to investigate auditory brain processing of musical sounds relevant to prosody processing in adolescent CI-users who have received their implant in childhood. Furthermore, we aimed to investigate the potential impact of intensive musical training on adolescent CI-users’ discrimination of music and speech prosody. Here we present preliminary analyses of ERP responses to rhythmically deviant stimuli and present results from a behavioral rhythm discrimination test. Eleven adolescent CI users (M.age = 17 years) participated in a group-based music training program, consisting of active music making supplemented with daily computer based listening exercises. NH participants received no training. Nine of the CI-users had bilateral implants and two had unilateral implants. The mean implant experience was 9.47 years (range: 1.8-15.2). Ten NH peers (M.age = 16.2 years) formed a reference group. Before and after the training, both groups underwent EEG recordings and behavioral tests for perception of music and speech. EEG was recorded with an adapted version of the musical multifeature paradigm (MuMuFe; Vuust, 2012), presenting a musical standard randomly violated by different musical deviants (pitch, timbre, rhythm and intensity). Difference waves for the rhythm deviant were analyzed in the time window between 300 and 320 ms. Separate mixed-model ANOVAs were performed for left and right fronto-central electrodes. For both left and right electrode sites we found a main effect of group (Left: F(1,16) = 39,859 p = 0.000, Right: F(1,16)=7.862 p = 0.013), driven by higher mean amplitude in the NH group. There was no main effect of training. Left hemisphere sites showed a significant group by session interaction (F(1,16)=11.211 p=0.004), driven by a larger difference wave (rhythm deviant - standard) in the CI group following training (t(7) = 4.05, p = 0.016 (Bonferroni corrected)). Right hemisphere sites showed no significant effect. The rhythm discrimination test showed a significant gain (t (1,11)=2.434; p=0.035) in the CI group after training. The NH group scored significantly higher than the CI group at both sessions. Our results suggest that adolescent CI users, who have only experienced sound through the implant, show brain responses to musical stimuli resembling those of NH peers, and that this response can be altered by intensive musical training. The finding points toward the possibility of improving appreciation of music in general for adolescent CI users, and using music as a motivating element in speech therapy programs.