The epidemiologic patterns of childhood hearing loss are ill-defined. Patterns found in prospective observations of newborns in intensive care nurseries are quite different from the patterns apparent in retrospective studies of hearing-impaired children. High-risk registers, at least those prior to the 1990 reformulation, missed about half of children subsequently identified as having hearing loss. Items comprising high-risk registers carry unequal weight. Children with sensorineural hearing loss have a disproportionately increased occurrence of otitis media, with its conductive hearing loss. Sensorineural hearing loss can be progressive. Case-control studies of children with hearing loss are needed. Efforts for the early diagnosis of hearing loss in children must take into practical consideration the uncertainties of available epidemiologic data, and the clinical nuances.