This article investigated to what extent age, use of a cochlear implant, parental hearing status, and use of sign in the home determine language of instruction for profoundly deaf children. Categorical data from 8,325 profoundly deaf students from the 2008 Annual Survey of Deaf and Hard-of-Hearing Children and Youth were analyzed using chi-square automated interaction detector, a stepwise analytic procedure that allows the assessment of higher order interactions among categorical variables. Results indicated that all characteristics were significantly related to classroom communication modality. Although younger and older students demonstrated a different distribution of communication modality, for both younger and older students, cochlear implantation had the greatest effect on differentiating students into communication modalities, yielding greater gains in the speech-only category for implanted students. For all subgroups defined by age and implantation status, the use of sign at home further segregated the sample into communication modality subgroups, reducing the likelihood of speech only and increasing the placement of students into signing classroom settings. Implications for future research in the field of deaf education are discussed.