OBJECTIVE: Childhood irritability predicts suicidal ideation/attempt (suicidality), but it is unclear whether irritability is an independent and direct risk factor for suicidality or a marker of intermediate mental health symptoms associated with suicidality. This study aimed to identify developmental patterns of childhood irritability and to test whether childhood irritability is directly associated with suicidality or indirectly associated with intermediate mental health symptoms. METHOD: One thousand three hundred ninety-three participants from the Quebec Longitudinal Study of Child Development were followed from birth to 17 years. Teachers assessed irritability yearly (at 6-12 years) and children self-reported intermediate mental health symptoms (depression, anxiety, disruptiveness, and hyperactivity-impulsivity; at 13 years) and suicidality (at 15 and 17 years). RESULTS: Four irritability trajectories were identified: low (74.7%), rising (13.0%), declining (7.4%), and persistent (5.0%). Children following a rising irritability trajectory (versus a low trajectory) were at higher suicidality risk. A large proportion of this association was direct (odds ratio 2.11, 95% CI 1.30-3.43) and a small proportion was indirect by depressive symptoms (accounting for 23% of the association; odds ratio 1.17, 95% CI 1.03-1.34). Children on a persistent irritability trajectory (versus a low trajectory) were at higher risk of suicidality and this association was uniquely indirect by depressive symptoms (accounting for 73% of the association; odds ratio 1.51, 95% CI 1.16-1.97). The declining trajectory was not related to suicidality; no association with anxiety, disruptiveness, and hyperactivity-impulsivity was found. CONCLUSION: Rising irritability across childhood represents a direct risk for suicidality. Persistent irritability appears to be a distal marker of suicidality acting through more proximal depressive symptoms.< Réduire