Search for Programs to Help YouthSearch for Programs to Help Youth

Aggression Replacement Training® (ART®)

Ages 12-17

Rating: Level 2


Aggression Replacement Training® (ART®) is a multimodal psychoeducational intervention designed to alter the behavior of chronically aggressive adolescents and young children. The goal of ART® is to improve social skill competence, anger control, and moral reasoning. The program incorporates three specific interventions: skill-streaming, anger-control training, and training in moral reasoning. Skill-streaming uses modeling, role-playing, performance feedback, and transfer training to teach prosocial skills. In anger-control training, participating youths must bring to each session one or more descriptions of recent anger-arousing experiences (hassles), and over the duration of the program they are trained in how to respond to their hassles. Training in moral reasoning is designed to enhance youths’ sense of fairness and justice regarding the needs and rights of others and to train youths to imagine the perspectives of others when they confront various moral problem situations.

The program consists of a 10-week, 30-hour intervention administered to groups of 8 to 12 juvenile offenders thrice weekly. The 10-week sequence is the “core” curriculum, though the ART® curriculum has been offered in a variety of lengths. During these 10 weeks, participating youths typically attend three 1-hour sessions per week, one session each of skill-streaming, anger-control training, and training in moral reasoning. The program relies on repetitive learning techniques to teach participants to control impulsiveness and anger and use more appropriate behaviors. In addition, guided group discussion is used to correct antisocial thinking. The ART® training manual presents program procedures and the curriculum in detail and is available in both English and Spanish editions. ART® has been implemented in school, delinquency, and mental health settings.


The ART® program has been evaluated in numerous studies. In general, the studies were comprehensive and used acceptable evaluation designs, psychometrics, and data analysis techniques. But many of the studies did not provide a demonstrated effect on violent behavior or on other conduct problems 1 year or longer beyond baseline.

One evaluation used a quasi-experimental design with nonequivalent comparison groups. The sample was collected from a New York State Division for Youth facility and included 60 youths, most of whom had been incarcerated for crimes such as burglary, robbery, and various drug offenses. Twenty-four of these youths received the 10-week ART® program. Another 24 youths were assigned to a no-ART®, brief-instructions control group. This condition controlled for the possibility that any apparent ART®-derived gains in skill performance were not due to ART® per se. Finally, 12 youths were placed in the no-treatment control group.

A second study was designed to both replicate the procedures and findings of the aforementioned study as well as extend them to youths incarcerated for substantially more serious felonies. The study sample included 51 youths who were incarcerated for murder, manslaughter, rape, sodomy, attempted murder, assault, and robbery. In all of its procedural and experimental details, the second study replicated the effort of the first. The second study employed the same preparatory activities, materials, ART® curriculum, testing, staff training, resident training, supervision, and data analysis procedures.

A third evaluation was designed to examine the efficacy of ART® as a community-based, postrelease intervention. This study also employed a quasi-experimental design with a three-way comparison of ART®. Condition 1 provided the ART® program to youths and to youths’ parents or other family members. Condition 2 provided the ART® program to youths only. Condition 3 provided neither parents nor youths with ART®. For the most part, youths were assigned to project conditions on a random basis, with departures from randomization becoming necessary on occasion as a function of the five-city, multisite, time-extended nature of the project.

A fourth study conducted by Washington State Institute for Public Policy used a pseudo–random assignment waitlist procedure to assign 1,229 adjudicated youths to either a control (n=525) or treatment group (n=704). Youths who met the selection criteria and had sufficient time on supervision to complete the program were assigned by court staff to the appropriate program. When the program reached capacity (all therapists had full caseloads or sessions were full), the remaining eligible youths were assigned by court staff to the control group and never participated in the program; instead, they received the usual juvenile court services. The sample was roughly 80 percent 15-year-old males. The analyses use multivariate statistical techniques to control for systemic differences between the program and control groups on key characteristics (gender, age, and domain risk and protective factor scores). Recidivism was measured by using conviction rates for subsequent juvenile or adult offenses. The follow-up “at risk” period for each youth is 18 months.


The findings from the first two studies reveal ART® to be an effective intervention for incarcerated juvenile delinquents. It enhanced prosocial skill competency and overt prosocial behavior, reduced the level of rated impulsiveness, and—in one of the two samples studied—decreased (where possible) the frequency and intensity of acting-out behaviors and enhanced the participants’ levels of moral reasoning.

The first study revealed that, compared with both control groups, youths who participated in the ART® program significantly acquired and transferred 4 of the 10 skill-streaming skills: expressing a complaint, preparing for a stressful conversation, responding to anger, and dealing with group pressure. Similarly significant ART®-versus-control-group comparisons emerged for the number and intensity of in-facility acting out and for staff-rated impulsiveness. During the 1-year follow-up, 54 youths were released from the facility. Of those released, 17 had received ART® and 37 had not. In four of the six areas rated—namely, home and family, peer, legal, and overall, but not school and work-ART®— youths were rated significantly superior at in-community functioning than were youths who had not received ART®. Similar findings were reported in the second study.

In the third evaluation (the postrelease community-based study), results indicated that, though they did not differ significantly from one another, the two ART® groups each increased significantly in their overall interpersonal skill competence compared with the control youths. Perhaps more important, however, rearrest rates were tracked during the 3 months in which youths in the two intervention groups received the ART® program and during the 3 subsequent no-ART® months. Meaningful differences in favor of the two intervention groups were found. Youths in both of the ART® groups were rearrested less often than youths not receiving ART®. And the ART® youths-plus-family-members group did better than the ART® youths-only group.

The Washington State study found that when ART is delivered competently, the program reduces felony recidivism and is cost effective. For the five courts rated as not competent, the adjusted 18-month felony recidivism rate is 27 percent compared with 25 percent for the control group. This difference is not statistically significant. However, for the 21 courts rated as either competent or highly competent, the 18-month felony recidivism rate is 19 percent. This is a 24 percent reduction in felony recidivism compared with the control group, which is statistically significant. Moreover, the cost–benefit analysis demonstrates that when ART is delivered by competent courts, it generates $11.66 in benefits (avoided crime costs) for each $1.00 spent on the program. When not competently delivered, ART costs the taxpayer $3.10. Averaging these results for all youths receiving ART, regardless of court competence, results in a net savings of $6.71 per $1.00 of costs.

Risk Factors


  • Anti-social behavior and alienation/Delinquent beliefs/General delinquency involvement/Drug dealing
  • Early onset of aggression and/or violence
  • Lack of guilt and empathy
  • Life stressors
  • Mental disorder/Mental health problem/Conduct disorder
  • Victimization and exposure to violence


  • Family history of the problem behavior/Parent criminality
  • Family management problems/Poor parental supervision and/or monitoring
  • Family violence
  • Pattern of high family conflict


  • Association with delinquent and/or aggressive peers

Protective Factors


  • Perception of social support from adults and peers
  • Positive / Resilient temperament
  • Self-efficacy
  • Social competencies and problem-solving skills


  • Effective parenting
  • Opportunities for prosocial family involvement


  • Presence and involvement of caring, supportive adults


  • Prosocial opportunities for participation / Availability of neighborhood resources
  • Rewards for prosocial community involvement


  • Good relationships with peers
  • Involvement with positive peer group activities


  • NIJ: What Works


Barnoski, Robert. 2004. Outcome Evaluation of Washington State’s Research-Based Programs for Juveniles. Olympia, Wash.: Washington State Institute for Public Policy.

Coleman, Maggie, Steven Pfeiffer, and Thomas Oakland. 1991. “Aggression Replacement Training With Behavior-Disordered Adolescents.” Unpublished manuscript. Austin, Texas: University of Texas, Special Education Department.

Curulla, Virginia L. 1990. “Aggression Replacement Training in the Community for Adult Learning-Disabled Offenders.” Unpublished manuscript. Seattle, Wash.: University of Washington, Special Education Department.

Glick, Barry. 1996. “Aggression Replacement Training in Children and Adolescents.” The Hatherleigh Guide to Child and Adolescent Therapy 5:191–226.

Goldstein, Arnold P., and Barry Glick. 1994. “Aggression Replacement Training: Curriculum and Evaluation.” Simulation and Gaming 25(1):9–26.

———. 1996a. “Aggression Replacement Training: Methods and Outcomes.” In Clive R. Hollin and Keith Howells (eds.). Clinical Approaches to Working With Offenders. Chichester, England: John Wiley & Sons.

———. 1996b. “Aggression Replacement Training: School-Based Instruction in Prosocial Skills.” The Quarterly Journal of the National Association of School Safety and Law Enforcement Officers.

———. 1996c. “Aggression Replacement Training: Teaching Prosocial Behaviors to Antisocial Youth.” In Robert R. Ross, Daniel H. Antonowicz, and Gurmeet K. Dhaliwal (eds.). Effective Delinquency Prevention and Offender Rehabilitation. Ottawa, Ontario: AIR Training and Publications.

Goldstein, Arnold P., Barry Glick, Wilma Carthan, and Douglas A. Blancero. 1994. The Prosocial Gang: Implementing Aggression Replacement Training. Thousand Oaks, Calif.: Sage Publications.

Goldstein, Arnold P., Barry Glick, and John C. Gibbs. 1998. Aggression Replacement Training: A Comprehensive Intervention for Aggressive Youth (revised ed.). Champaign, Ill: Research Press.

Goldstein, Arnold P., Barry Glick, M.J. Irwin, Claudia Pask–McCartney, and I. Rubama. 1989. Reducing Delinquency: Intervention in the Community. New York, N.Y.: Pergamon.

Goldstein, Arnold P., Barry Glick, Scott Reiner, Deborah Zimmerman, and Thomas M. Coultry. 1987. Aggression Replacement Training. Champaign, Ill.: Research Press.


Barry Glick, Ph.D.
G&G Consultants, LLC
106 Acorn Drive, Suite A
Glenville, NY 12302
Phone: (518) 399-7933
Fax: (518) 384-2070
Web site: