ABSTRACT: Bedside courses are of outstanding importance when training medical students. The fact that less and less teaching is taking place nowadays at the patient's bedside makes it all the more important that the available time be put to effective use. The aim of this study was to check whether structured improvement measures in the course (scripts, lecturer briefing, e-learning cases) would improve the abilities of the students on the basis of a subjective self-assessment as well as an external assessment by the lecturers with respect to clinical abilities.Bedside teaching takes place in the fourth study year in the Medical Clinics of the TU Munich. Both students and lecturers had the chance to hand in an anonymous, quantitative self- and external assessment of the clinical abilities of the students (German grading system) after every course date. This assessment took place online in the three categories "Medical history & examination", "Diagnosis" and "Therapy". An overall period of four semesters, each with 6 course dates, was investigated. After two of the total of four semesters in the study, the course was changed by introducing scripts, lecturer briefing as well as interactive e-learning cases. The self- and external assessment was compared both within the semester (date 1-3: A; date 4-6: B), during the course as well as before and after introducing the improvement measures ("before" (T0): SS 2012, SS 2013, "after" (T1): WS 2013/2014, SS 2014).There was a significant improvement in one's own abilities on the basis of the self-assessment within each semester when comparing the first (A) and the last (B) course dates. Moreover, there was a significant improvement in the performances in all three categories when T0 was compared with T1, from both the point of view of the students ("Medical history & examination": T0 =2.5±0.9, T1=2.2±0.7, pp<0.001; "Diagnosis" T0=3.1±1.0, T1=2.8 ±0.9, pp<0.001; "Therapy": T0=3.8±1.3, T1=3.5±1.2, pp<0.018) and in two of the three categories from the point of view of the lecturers ("Diagnosis": T0=3.0±1.0, T1=2.7±0.7, p.=0.028; "Therapy": T0=3.8±1.1, T1=3.1±1.0, p<0.001).The structured measures to improve the course including the interactive e-learning cases could have contributed to improved practical abilities with respect to the medical history and examination techniques as well as diagnostic and therapeutic thinking. The external evaluation by lecturers confirmed the improvement with respect to the diagnostic and therapeutic abilities. They only saw no dynamic change in the student's taking histories and clinical examinations.