Student characteristics associated with school failure and school dropouts are well documented. This study sought to determine if a subgroup of students could be differentiated prior to high school as at highest risk for secondary school failure and dropout. It also attempted to describe behavioral, academic, and familial differences and similarities between the hghest-risk students and their at-risk peers during the first 2 years of junior high school. Subjects (N=351) were sixth-graders who were rated by classroom teachers on the need for supervision; level of motivation; academic potential; social interaction skill; and teachability. Based on results, 78 students were classified as highest risk (HstR) and 273 were classified as at-risk (AR); the groups were reduced in size but retained the statistical characteristics of the original pools; i.e., HstR was reduced to 24 students and AR to 29. In terms of junior high school adjustment, the HstR students could be distinguished from the AR students on intensity of unacceptable behavior. Results showed that AR students received significantly better report card grades than HstR students in academics as well as in work habits and cooperation. The HstR students were not absent more often than the AR students. In terms of predicting eighth grade adjustment, the teacher scale accounted for a statistically significant proportion of eighth grade variance for both groups in selective truancy, periods absent, report card grades, work habit grades, and cooperation grades. Results lend strong initial support to pursuing research efforts testing the efficacy of targeting HstR students on the five item characteristics. (ABL)