Background Previous studies have noted that a past suicide attempt (SA) predicts a future SA, but few studies have reported whether previous SAs that predict a future attempt differ from those that do not. Knowing which characteristics of previous SAs predict future attempts would assist in evaluating adolescents at risk of attempt repetition. This longitudinal study of an unreferred sample examined which characteristics of adolescent SAs increased risk for repeat attempts. Methods Fifty‐four adolescents who had attempted suicide were identified through a two‐stage screening of 1,729 high school students. Adolescents reported details of their past SA on the Adolescent Suicide Interview and were reassessed 4–6 years later by telephone. Results Eighteen of the 54 teens (33%) reported that they had made another SA since baseline, and 17 of these reported characteristics of their later attempt. The odds of a further attempt were significantly increased by being alone (OR = 6.1, 95% CI = 1.1–34.8), retrospectively reporting a serious wish to die (OR = 5.2, 95% CI = 1.2–22.7), and planning the attempt for an hour or more (OR = 5.1, 95% CI = 1.1–25.0). The method of attempt remained consistent from baseline to follow‐up attempt (κ = .67). Conclusions Screening high school students to identify those who are at risk for making future SAs should include questions about number of previous SAs and such indicators of risk as isolation, wish to die, and extent of planning prior to a SA.