Purpose: to verify the hospital discharge rate in breastfeeding and speech-language practice during implementation of the kangaroo method in a tertiary-level teaching hospital. Methods: a retrospective and documentary study, whose data were collected from electronic medical records of all preterm and/or low birth weight newborns, born at or admitted to a teaching hospital, from August 2016 to August 2017, and who had received care by the speech-language team. Data were collected regarding sample characterization (gender, Apgar, gestational age, birth weight, and orotracheal intubation time); speech-language therapy time; description of the procedures used, such as weight to start breastfeeding, empty breast technique, translactation, and feeding method at the time of speech-language therapy and hospital discharge. Appropriate statistical tests were applied, adopting a significance level lower than 0.05%. Results: there was a higher rate of breastfeeding in moderately preterm infants, born in a gestational age of 30 to 34 weeks, and in low birth weight newborns. Conclusion: breastfeeding was the prevalent feeding method for newborns. The results show that the speech-language stimulation of newborns was adapted from the conventional model to that proposed by the Kangaroo Method, corroborating with humanized care guidelines.