While grief is a universal experience, this phenomenon is experienced in a variety of ways largely dependent upon one's cultural schema. To consider the potential problems inherent in generalizing the results of an assessment tool across cultures, this study explores the notion of generalizability in assessment by evaluating the reliability of the Texas Revised Inventory of Grief (TRIG) using a meta-analytic technique called reliability generalization. The TRIG demonstrated strong reliability with mean Cronbach's alphas of .90 and .82 for the present and past subscales, respectively. Overall, the TRIG, especially the present subscale, appears to produce reliable scores even across cultures.