This article proposes a new measurement instrument of trauma-related shame. The purpose of this study is to investigate the psychometric properties of the scores derived from the Trauma Related Shame Inventory (TRSI) by means of generalizability theory (G-theory). The psychometric analyses are based on a sample of 50 patients in treatment for Post-traumatic Stress Disorder (PTSD). The results provided supporting construct validity evidence for the interpretation of TRSI as a homogeneous construct. The 24-item version of internal and external referenced shame yielded generalizability and dependability coefficients of .874 and .868, respectively. The distinction between shame and guilt was supported by a high generalizability coefficient of .812 for the difference scores between TRSI and guilt cognition scale. Further validity evidence was provided by a positive relationship between TRSI and a) self-judgment subscale in Self-Compassion Scale (SCS; Neff
Self and Identity 2:(3), 223–250,
2003) and b) Beck Depression Inventory (Beck Steer and Brown
1996a) when controlled for guilt. The results of the present study provided promising support for using the 24-item version of TRSI in both clinical research and practice.