Research on religiosity and aging has focused more on church attendance than on the religious role and the strength of beliefs and the extent of private devotions. This study of 123 older West Texans examines the degree to which nine variables account for variation in both organized religious activities and private religious behavior. These include three social activity and interaction variables, three religion variables, and three personal variables (health, age, and income). Strength of religious conviction proved to be the strongest predictor of both types of religious participation. Multiple-regression analysis did not support the hypothesis that private religious activity is compensatory; that is, it does not make up for increased social or personal deprivation. Poor health, low income, reduced activity, and living alone did not predict higher levels of nonorganized religious behavior. However, strong kin/friend networks did predict high levels of private devotion. This research suggests that frequent interaction in a social network contributes to the spiritual well-being of elderly persons; that is, it affirms the wholeness of their lives.