Cochlear implants (CIs) are the most successful sensory implant in history, restoring the sensation of sound to thousands of persons who have severe to profound hearing loss. Implants do not recreate acoustic sound as most of us know it, but they instead convey a rough representation of the temporal envelope of signals. This sparse signal, derived from the envelopes of narrowband frequency filters, is sufficient for enabling speech understanding in quiet environments for those who lose hearing as adults and is enough for most children to develop spoken language skills. The variability between users is huge, however, and is only partially understood. CIs provide acoustic information that is sufficient for the recognition of some aspects of spoken language, especially information that can be conveyed by temporal patterns, such as syllable timing, consonant voicing, and manner of articulation. They are insufficient for conveying pitch cues and separating speech from noise. There is a great need for improving our understanding of functional outcomes of CI success beyond measuring percent correct for word and sentence recognitions. Moreover, greater understanding of the variability experienced by children, especially children and families from various social and cultural backgrounds, is of paramount importance. Future developments will no doubt expand the use of this remarkable device.