Scientists have been exploring the field of Readability for many decades through various tools, such as special software or indicators (Gulpease, Dale-Chall, Gunning Fox, etc.), in order to make careful measurements of the Readability degree. The purpose of the present research is to identify those criteria, which first determine the level of difficulty, and second, the language level of a written text. All data were collected from the examinations of the Greek Certificate (KPG) of May 2015 and November 2016. 316 written texts, including both levels, B1 and B2, were digitized manually in Word form. In the second phase, they were measured by using the Read-It tool, and the values and results of this process were evaluated. Through SPSS.24, and in particular, with factor analysis, the final results were achieved. Summarized, inter alia, the following were found: 1. The texts produced by Greek users of the Italian language show that the Readability degree seems to vary by language level and degree of difficulty both through the use of vocabulary as well as grammatical and syntactic features. 2. There seems to be a great difficulty in producing secondary texts, correctly worded in Italian, at both levels. 3. Another phenomenon that reduces both, the readability grade and the languagelevel is the vocabulary confusion with other languages, e.g. we see Greek users writing English or Spanish words confusing them with Italian. These valuable findings and this research are likely to pave the way for future scientists to delve even deeper into the parameters of writing with the ultimate goal of developing more advanced software that will help to improve the use of languages by foreign users and to prepare tests more accurate and fair by certification entities.