It has been well documented that a history of suicide attempts confers risk for subsequent attempts; however, efforts to explain how variables may change after a previous attempt and in turn relate to future suicide attempts are rare in the literature. This study presents longitudinal data on adolescent suicide attempts in Hong Kong, and examines whether the data support the "crescendo" model to explain repeat suicide attempts. One thousand and ninety-nine community adolescents aged 12-18 years were evaluated at two assessment points 12 months apart (T1 and T2). The study assessed (1) risk factors at T1 for a suicide attempt between T1 and T2, (2) whether a suicide attempt during the 12 months prior to T1 predicted an attempt between the two assessment points, and (3) whether the indicators of distress worsened from T1 to T2 if an attempt had taken place in the interim. The results indicated that: (1) depressive symptoms, substance use, and suicidal ideation measured at T1 were independent predictors of a suicide attempt between T1 and T2; (2) suicide attempt in the year prior to T1 predicted suicide attempt between T1 and T2 after controlling for other predictors; and (3) suicide attempt between T1 and T2 was a predictive factor for a negative change from T1 to T2 in substance use, suicidal ideation, family relationships, depression, anxiety, and life stress. These findings are consistent with the "crescendo" model proposing that the risk of repeat attempts is enhanced following a previous suicide attempt.