The CEFR is increasingly being used as the framework of choice for the assessment of language proficiency at universities across Europe. However, to attain consistent assessment, familiarization and standardization are essential. In this paper we report a case study of embedding a standardization procedure in writing assessment activities at the University of Groningen. The project shows the value of standardization procedures within the CEFR and reports on the difficulty of finding consistently assessed ‘flat’ samples. Moreover, it reports on the desirability to scaffold teacher training sessions using flat samples with less ‘even’ samples.