Abstract We investigated (a) whether psychosocial factors (experienced stress, anticipatory worry, social detachment, sleeping disturbances, alcohol use) predict the course of paranoid ideation between the ages of 24 to 50 years and (b) whether the predictive relationships are more likely to proceed from the psychosocial factors to paranoid ideation, or vice versa. The participants (N = 1534–1553) came from the population-based Young Finns study. Paranoid ideation and psychosocial factors were assessed by reliable self-report questionnaires in 2001, 2007, and 2011/2012. The data were analyzed using growth curve and structural equation models. High experienced stress, anticipatory worry, social detachment, frequent sleeping disturbances, and frequent alcohol use predicted more paranoid ideation. More risk factors predicted increasing paranoid ideation. There were bidirectional predictive relationships of paranoid ideation with experienced stress, anticipatory worry, social detachment, and sleeping disturbances. The link between alcohol use and paranoid ideation was only correlative. In conclusion, paranoid ideation increases by reciprocal interactions with stress, worry, social detachment, and sleeping disturbances. The findings support the threat–anticipation model of paranoid ideation, providing important implications for treatment of paranoia.