Using the technique of meta-analysis, the four authors reviewed 643 studies collected from a computer-based search of Psychological Abstracts. Search words were 'juvenile delinquents,' 'treatment,' 'intervention,' and 'outcome studies.' Of the 643 studies, 163 appeared to be outcome studies with juvenile delinquents. Of those 163 studies, 91 were codable for the meta-analysis. Of the 91 codable studies, 25 were concerned with behavioral interventions. The procedures used required that all research designs be reduced to comparisons between two groups or two periods, resulting in a meta-analysis of 30 distinct coded studies. The study characteristics that were coded and the indexes used in the calculation of effect size in the meta-analysis are discussed in detail. The descriptive findings of the meta-analysis include study characteristics, investigator characteristics, subject dimensions, intervention components, and methodological characteristics. While the researchers' analysis of the effectiveness of the interventions studied matched those of the studies' authors and reported consistently positive results, the calculated effectiveness method as used in the meta-analysis yielded less than positive results. The authors point out that the meta-analysis review indicates that outcome studies should be conducted with more methodological vigor. The long-term impact of behavioral interventions with delinquents will eventually be measured by better quality research carried out in settings closer to the followup environment. 61 references.