This article models the migration flow of EEA students who graduated from masters and doctoral programmes in UK universities. The increased intra-European mobility of students and graduates is claimed to have crucial positive influence on building Europe's high-skill labour force, which in turn would strengthen its competency in the global knowledge economy. However, the absence of accurate quantitative data on degree-mobile students makes it difficult to track and investigate their post-study mobility patterns. Six one-year extracts from the DLHE dataset were analysed (2011/2012-2016/2017) using cross-classified multilevel modelling in order to investigate which group of EEA graduates were more likely to remain and work in UK, and how the patterns were changing in the long-term. The stay rates of students from four countries (Romania, Bulgaria, Italy, and Greece) were found to have stable growth between 2011/2012 and 2016/2017. Graduates who received degrees in Medicine & Dentistry and Computer Sciences were found to have the highest and the most stable stay rates among all graduates. Multilevel modelling results show that, at domicile-level, the difference between home and host country in GDP per capita could be an effective predictor in analysing student post-graduation movement. At HEI-level, the prestige level of HEI could not effectively predict students' stay rate. At individual-level, students' education background and gender identity were significant in predicting stay rate. Results imply that group effects at neither HEI-level nor domicile-level should be ignored. This study provides an empirical foundation for evidence-based decision-making in a field that is heavily contested in policymaking.