Creative activities in music promote the ability for self- expression, stimulate children's imagination and contribute to their active involvement into the learning process. Musical creativity in the school can take place in different kinds of activities described as listening, performing, improvisation and composition. The two latter have been the area of interest for most studies conducted on children's musical creativity over the past few decades, investigating their original music making, in different settings, within different frameworks and with children using their voices, their bodies or musical instruments to make music. However, research on the assessment of creative thinking is not at all extended and whether we can assess creativity or not has been a controversial issue. This paper is based on a study that investigated the role of improvisation in the development of children's creative thinking in music. In particular, the study addressed the following questions: Are there any differences between the creative products of children engaged in an intervention improvisation programme and those who are not? If so, to what extent are various dimensions of creative thinking promoted, namely musical extensiveness, flexibility, originality and syntax? A pre-test post-test quasi- experiment (pre-test post-test randomised controlled trial) was designed and two groups of children took part; one of them enrolled in an improvisation-based training programme during the weekly music lessons. Webster's Measure of Creative Thinking in Music (MCTM-II) was used as a basis for assessing the creative thinking for such factors as musical extensiveness (ME), flexibility (MF), originality (MO) and syntax (MS). The analysis showed that the two groups (control and experimental) results in the post-test differed, with the experimental group scoring higher results (F = 42.515, mean square = 8.908, p< .000), which shows that improvisation (independent variable) had a effect on the experimental group's results. The experimental group scored significantly higher in some particular factors: MF (F = 48.571, mean square = 13.281, p< .000), MO (F = 43.196, mean square = 14.626, p< .000), MS F = 47.603, mean square = 13.472, p< .000). However, some problems might occur when assessing children's musical creativity. In the particular study, these problems were mainly relevant to developmental issues, such as the ability of the children to use musical patterns, and the issue of familiarity to concepts like 'low and high'. Some other problems also could emerge from an attempt to measure creativity, like the level of freedom that a test allows to children, the time limitations, the choice of instruments, and the person who conducts the test. Further research could shed light on the above issues and could also examine various factors that might interfere with the assessment of creative thinking, such as the music tradition of each school, the teachers' qualifications and skills and the children's social-cultural background.