Violent incidents occur frequently in schools and suicide ranks second to accidents as the leading cause of death among adolescents. The purpose of this presentation is to summarize the findings of six studies that used a stimulus-drawing task for access to fantasies, thoughts, and feelings. The studies were based on the theory that the task can serve as a first step in identifying children and adolescents at risk. The studies also examined gender and cultural differences or similarities, and were conducted in schools or in facilities for delinquent adolescents in the U.S.A, Russia, and Thailand. The task presents line drawings of people, animals, places and things, asks respondents to choose two, imagine something happen between them, draw what they imagine, and add titles or stories. Responses are scored on rating scales that range from strongly negative, such as suicidal or homicidal themes to strongly positive, such as loving relationship. In the first study, the task was presented to 290 children, adolescents and adults; 50 were clinically depressed; the others were normal or had different disabilities. Results indicated that significantly more depressed subjects scored 1 point than any other group. Inter-scorer and retest studies found the scale reliable. The second study asked if self-images in the responses of 64 delinquent adolescents could be identified without talking to those who drew them. The third study examined responses by 64 delinquent and 74 non-delinquent adolescents for attitudes toward self and other, finding no significant differences in gender or delinquency, but significant differences in assaultive and solitary content as well as differences in gender and delinquency. The fourth study used two scales in comparing responses by 30 students who had histories of aggressive behavior with 181 non-aggressive students. One scale assessed responses for emotional content; the other, assessed self-images. The study found significant differences between the groups of aggressive and non-aggressive students. Gender differences and two subgroups of aggression also emerged: reactive aggression and predatory aggression. In the fifth study, Russian investigators compared 27 delinquent and 25 non-delinquent adolescents, finding no significant cultural differences in subjects scoring 1 point on both scales but significant differences in self-image scores and between experimental and control groups. The sixth study, in Thailand, assessed effects of an art therapy program on delinquent adolescents, dividing the sample into experimental and control groups. Pre- and post-test scores of the experimental group and control group differed significantly. The studies seem to support the theory, and suggest that respondents who score 1 point on both scales, or else 1 point in emotional content combined with 5 points in self-image be promptly referred to mental health professionals for further evaluation.