Objective: Previous research has shown contradictory evidence for the relationship between religiosity and trauma; exposure to traumatic life events has been associated with both increases and decreases in religiosity over time. On the basis of a long theoretical tradition of linking death and religious belief and recent empirical evidence that thoughts of death may increase religiosity, we tested whether one determinant of trauma’s influence on religion is the degree to which it makes death salient. Method: Using longitudinal data from the Dunedin Multidisciplinary Health and Development Study, a unique population-representative birth cohort, we tested whether the relationship between trauma and religiosity depends on whether the trauma involves death. Participants reported their private, ceremonial, and public religious behaviors at ages 26 and 32 and, at age 32, whether they had experienced any of 23 traumatic life events since age 26. Results: Experiencing the death of a loved one (but not an equally traumatic event not involving death) predicted a future increase in private religious behavior (e.g., prayer) among those already practicing such behaviors, and an increase in the importance of religious ceremonies among those with relatively little prior interest in them. On the other hand, experiencing a death-unrelated trauma predicted a future reduction in public displays of religiosity among those previously so inclined. Conclusion: The study represents a significant step in understanding religious responses to trauma, and emphasizes the importance of considering not only the nature of a trauma, but also the dimensions and practices of a victim’s religiosity prior to it.