The classical languages have been affected by the curricular reform in Croatia. The new Ancient Greek language curriculum sets text comprehension as its main goal, which means that students should be better prepared for reading original texts. In that context, we want to study the relation between the real state of the Ancient Greek, as attested by its texts, and its teaching in schools. In this paper we examine the possibility of creating a curriculum for Ancient Greek based on corpus frequency analysis. We first examine whether the current curriculum, introduced in 2019, is based on the frequency of forms in Ancient Greek or not. In other words, we are trying to see if it prepares the students to read texts by favouring those linguistic forms that occur more frequently. We have determined that the current curriculum is not based on frequency of forms, but rather on traditional grammars and that it groups together forms with similar word-formation or inflection rules, teaching first what is regarded as simplest. Then we analyse a corpus of texts from The Reader for Greek Morphology 1, prepared at the University of Zagreb, Faculty of Humanities and Social Sciences. We describe the structure, extraction and the statistical analysis of the data and compare the results with those of similar studies (Mahoney, 2004 and Major, 2008). In the end, we describe the main features of a frequency-based curriculum according to the results of our analysis.