The demand for palliative care is increasing worldwide. Beyond the acquisition of technical knowledge, the development of adequate personal disposition toward the relationship with the dying is a key aspect of the future training of doctors. This study evaluates the psychometric properties of the 9-Item Version of the Frommelt Attitude Toward the Care of the Dying Scale (FATCOD-9IT) the authors recently developed and its capability to distinguish medical students with different attitudes toward the care of the dying and at different stages of medical training. The study included 595 medical students, 400 at the first and 195 at the fifth year. The Rasch rating scale model was specified to assess scale dimensionality, functioning and measurement invariance. Internal consistency, test-retest reliability and between-group difference sensitivity (first-vs fifth-year students) were evaluated using Cronbach's alpha, Intraclass correlation coefficients, Paired sample t-test and Mann-Whitney U. Scale unidimensionality, rating scale functioning and measurement invariance were established. The scale demonstrated good internal consistency and test-retest reliability, and adequately discriminated between first- and fifth-year students. The study supports the validity and reliability of the FATCOD-9IT. Its effectiveness, simplicity of compilation and score calculation, and gratuitousness encourage its widespread use as fast assessment of the medical student attitudes toward the care of the dying.