The relationship between spirituality and depressive symptoms among the Chinese elderly is not well known. The current study explores this relationship using longitudinal data and trajectory modeling of depressive symptoms. A longitudinal study design was used to measure depressive symptoms repeatedly from 2012 to 2021 using the Geriatric Depression Scale (GDS). Group-based trajectory modeling analysis was conducted to determine the trajectories of depressive symptoms, and multiple logistic regression was used to explore the association between spirituality and depressive symptom trajectories. A total of 2333 participants completed at least two GDS measures, and these were included in the Group-based trajectory modeling analysis. An optimal model of three trajectories was derived: no depressive symptoms group (75.2%), new-onset depressive symptoms group (14.4%), and persistent depressive symptoms group (10.4%). Logistic regression modeling revealed that higher spirituality was associated with a lower risk of both new-onset depressive symptoms (OR=0.68, 95% CI=0.49-0.93) and persistent depressive symptoms (OR=0.32, 95% CI=0.23-0.45). Spirituality predicts a lower risk of new-onset depressive symptoms and persistent symptoms among older adults in mainland China.