PURPOSE: to check the reliability of the information provided by adults and children about the habitual tongue position. METHODS: we investigated 30 children and 30 adults in two phases, with minimum difference of seven and maximum of twenty-one days. Initially we observed the usual position of the tongue. Then the subjects were questioned about the habitual position After the response, the tongue was stimulated with a wooden spatula, in order to enhance perception. Then, the subjects were guided to observe where their tongue was usually positioned in the oral cavity, until the second assessment. This time, the subjects were asked about their habitual position of the tongue. Data were analyzed using Kappa statistic. RESULTS: it was not possible to observe the usual position of the tongue in 100% of the sample. As for the general reliability of the responses, it was found between mild and regular classification. The children showed responses to be little consistent and very diverse. As for the adults, part of them submitted correct answers in the first question and others only submitted reliable answers after intra-oral perception stimulation. CONCLUSIONS: the reliability of the information provided by the individuals in the sample on the usual position of the tongue varies from mild, regular, and therefore low, both in children as in adults. A possible strategy to be used in clinical speech therapy practice is questioning the patients about their tongue position after a period of observation.