The most commonly used methods for analyzing change in crime and delinquency rates, longitudinal and cross-sectional techniques, offer incomplete and inadequate explanations. Cohort analysis, however, incorporates these two techniques into a delineation of age, period, and cohort effects on changing deviant activity. A brief illustration of the method is presented using male and female personal offense rates for three birth cohorts. A comparison of results from longitudinal, cross-sectional, and cohort analysis shows the improvement gained by utilizing the cohort technique. 2 tables, 5 footnotes, and 19 references. (Author abstract modified)