The 1974 sample was composed of 470 adjudicated boys from the Montreal area. They were aged from 13 to 18 and were recruited at the juvenile court. They were adjudicated under the Canadian Juvenile Delinquents Act or under the Quebec Youth Protection Act. Two years later, 396 of the 470 boys were reinterviewed. The 1992 sample involved 506 adjudicated boys from 12 to 18 years old. They lived in the Montreal area, and they were adjudicated under the Canadian Young Offenders Act or under the Quebec Youth Protection Act. The boys of the two samples were placed in institutions of various security levels or on probation because of a criminal act or a serious behavioral problem. Because the age distributions differed in the two samples, two subsamples were compared. The results for both age groups and both decades support the presence of a construct of general deviance. The analyses were conducted with a common set of measures of six types of deviance. This study reported variances between 48 percent and 54 percent. These proportions are comparable to the proportions that are reported in the literature. Since the proportions of explained variance were comparable and there were no substantial changes in the factor loadings, notwithstanding that the relative position of the types of deviance changed, the study concludes that the structure of deviance was stable between the 70s and the 90s. 3 tables, 1 note, and 21 references