The paper presents the first known longitudinal study of the relationship between loneliness, depressive symptoms, and suicide ideation in adolescence, in a stratified sample of high school students (Time 1 N = 1009; 57 % female; Time 2 N = 541; 60 % female). Cross-lagged structural equation modeling indicated that depressive symptoms led to more loneliness across time, whereas loneliness did not predict higher levels of depressive symptoms across time. Loneliness was found to be a correlate of depressive symptoms at the cross-sectional level, independent of gender, other demographic factors, multiple psychosocial variables, and social desirability. Loneliness did not predict suicide ideation over time or at the cross-sectional level, when controlling for depressive symptoms. Gender did not predict loneliness, depressive symptoms or suicide ideation across time. Future longitudinal studies of the relationship between loneliness, depressive symptoms, and suicide ideation in adolescence should use more extensive designs.