Loneliness is common among young people and is associated with negative health outcomes. Because loneliness is associated with a bias for interpreting social situations as threatening, cognitive bias modification for interpretation (CBM-I) training is a potential early intervention tool. We developed and delivered a single session of mental imagery enhanced digital CBM-I training, assessing feasibility, acceptability, and magnitude of change in interpretational style and loneliness.CBM-I training materials were developed using a co-creation approach with 18-25-year-olds with experience of loneliness. Another group of 18-25-year-olds with high loneliness received either online CBM-I (n = 29) or control (n = 27) training.CBM-I training uptake and retention rates were 88% and 92%, respectively. Participants found the training acceptable. The CBM-I group showed a reduction in social threat interpretations (d = 0.77), an increase in social benign interpretations (d = 0.84), and a decrease in loneliness (d = 0.56). The control group showed a small reduction in social threat interpretations (d = 0.21), no change in social benign interpretations (d = 0.04), and an increase in loneliness (d = 0.41).Interpretation biases relevant to youth loneliness may be modifiable, and CBM-I training could reduce feelings of loneliness. This informs psychological models of loneliness, and the development of CBM-I interventions targeting loneliness in young people.