The Radboud University Medical Center has a problem-based, learner-oriented, horizontally, and vertically integrated medical curriculum. Anatomists and clinicians have noticed students' decreasing anatomical knowledge and the disability to apply knowledge in diagnostic reasoning and problem solving. In a longitudinal cohort, the retention of anatomical knowledge gained during the first year of medical school among second-year medical students was assessed. In May 2011, 346 medical students applied for the second-year gastro-intestinal (GI) tract course. The students were asked to participate in a reexamination of a selection of anatomical questions of an examination from October 2009. The examination consisted of a clinical anatomy case scenario and two computed tomography (CT) images of thorax and abdomen in an extended matching format. A total of 165 students were included for analysis. In 2011, students scored significantly lower for the anatomy examination compared to 2009 with a decline in overall examination score of 14.7% (±11.7%). Decrease in knowledge was higher in the radiological questions, compared to the clinical anatomy cases 17.5% (±13.6%) vs. 7.9% (±10.0%), respectively, d = 5.17. In both years, male students scored slightly better compared to female students, and decline of knowledge seems somewhat lower in male students (13.1% (±11.1%) vs. 15.5% (±12.0%), respectively), d = -0.21. Anatomical knowledge in the problem-oriented horizontal and vertical integrated medical curriculum, declined by approximately 15% 1.5 year after the initial anatomy course. The loss of knowledge in the present study is relative small compared to previous studies.