Tonal Music is the focus of most of the current approaches to Music Education. For the last several decades, research about the way in which the tonal system is experienced by the listener has been developed. Because of this, it is now possible to find models which are able to explain such processes. Some of these come from the field of experimental research (Krumhansl, 1990; Bharucha, 1984; Butler & Brown, 1994; Dowling, 1994). Others belong to the field of music theory and analyse the processes experienced by an ideal listener and, whether explicitly or implicitly, state basic assumptions related to the cognitive processes of that listener. The Psychology of Music has begun to test such models empirically. Yet, in spite of being important for musical education, few studies focus on the analysis of these processes from a developmental perspective. Hence, it is difficult to have a complete understanding of the central role of tonality in the development of Music Education (see Lamont, 1998). The present study is based on the theoretical principles of H. Schenker ([1935]-1979). One of the most important cognitive assumptions of this theory is related to the concepts of structure and prolongation (Salzer, [1962]-1990; Forte & Gilbert, [1982]-1992). Some aspects of the theory have been tested experimentally (Serafine, Glassman & Overbeeke, 1989) providing some evidence for the cognitive reality of Schenker’s model. However, these studies tested very complex aspects of this theory. The purpose of this study has been to explore evidence about the acquisition of structural hearing (Salzer [1962]-1990) from a developmental perspective. Results of an experiment with children between six and fourteen years old are presented. A test of similarity judgement was done in which children had to judge pairs of melodies. The first melody of each pair was a model selected from the western academic repertoire. The second melody, a foil which was composed, was one of two types: either a) one which was identical in its underlying structure to the model melody and differed in the surface other melody was a foil that was identical in its underlying structure and differed only in surface; or b) one which was different from the model melody both in its underlying structure and its surface. The surface differences of type a) foil and type b) foil were composed in such a way that they were similar. Findings of Serafine & others (1989) provide evidence that an implicit recognition of structure would occur and that this recognition would be present in this type of similarity judgement tasks. Thus, subjects faced with a similar amount of surface differences, would be able to identify the melody that shares the underlying structure as being more similar to the model. In this exploratory study we have hypothesised that some of the variables involved in structural hearing are: the quality of the surface to convey direction and its relationship with the conveyed direction of the underlying structure, the actual length of the prolongation. In this way there would be certain melodic movements that would benefit the prolongation of the underlying structure. In addition, it has been hypothesised that, as long as tonality is acquired by cultural exposure (Sloboda, 1985) the ability to implicitly recognise the components of structure and prolongation would be dependent on age.