there is a great difficulty in determining earlier on which children would benefit or not from cochlear implants, especially because of their young age, the responses they give are very subtle. aim: to compare results obtained through video-recording of the interactions of children who may receive a cochlear implant with the results obtained through evaluation protocols. method: seven children, with an average age of 39.7 months, with profound hearing loss were selected for the study. it-mais and muss questionnaires were given to their parents/guardians of these children and the results were compared with the observation of the video-recordings. results: it was possible to observe that the data is compatible with the auditory stages. however, the muss questionnaire data gathered during playful activities is very different . the questionnaire only takes into consideration the use of verbal language and therefore the majority of the evaluated children inevitably score low. conclusion: observing children play allows us to trace a better profile of linguistic behavior and aspects relative to language, that may presented differences in the questionnaire.