Objective The main objective of this doctoral research is to get an insight into the change of students' expectations and assessments of their mentors' competence. We have also dealt with the self-estimations of mentors at different study years working with the same generation of students throughout the study. Subjects and methods The research was a longitudinal. cross-sectional and prospective one. It was conducted during a period of three academic years at the Clinical Hospital (KBC) Osijek. in the clinics and departments - teaching bases of the Osijek Faculty of Medicine. The study involved two groups of participants: three generations of students (2011/2012. 2012/2013. 2013/2014) of the first. second and third year at the University undergraduate study of Nursing at the Osijek Faculty of Medicine. as well as the clinical practise mentors for three courses. The testing instrument was a questionnaire on general characteristics of the respondents and 52 attributes of mentors’ competences divided into six categories. The instrument was adapted from the Nursing Clinical Teacher Effectiveness Inventory (NCTE). Criteria for inclusion of courses in each study year were: the course of health care. at least 60 hours of clinical practise. at least 8 ECTS credits. at least eight mentors. Results The mentors' competences were estimated at a significantly lower rate than expected by students of the first and the third year of study at all three academic years (p<0.001). The mentors’ self-estimated competences in performing their clinical practice were significantly higher by the mentors of all three study and academic years. in relation to the assessment of students (p<0.001). In the academic years of 2011/2012 and 2013/2014, students of the second year’s expectations were significantly lowest (p<0.001). while the lowest assessments were those of the first year students (p<0.001). In the academic year of 2012/2013. the lowest expectations and evaluations were those of the first year’s students (p<0.001). Mentors at the third year of 178 Robert Lovric Disertaciia. 2015. study in all three academic years significantly overestimated their own competences (p<0.001). Students at the same study years. but in the different academic years. had significantly different expectations (p<0.001) and estimates of mentors’ competence (p<0.001). Among the repeated mentors’ self-estimations. there were no significant differences at the first (p=0.584), the second (p=0.587) and the third (p=0.292) year of study. Expectations of students during the three-year study significantly increased every year (p<0.001). During their studies. the students had estimated their mentors’ competence as highest at the second year of study (p<0.001). Conclusion Students' expectations and estimations of mentors’ competences, as well as mentors’ self-assessments. differ with given study year. Students’ expectations grow with each study year. The students' assessment of mentors’ competence primarily depends on the mentors and the competences they apply in clinical practice.