Data were obtained from a nationally representative sample (n=893) of 14- to 25-year-olds in England and Wales, with a secondary sample (n=823) of 14- to 25-year-olds in high crime areas and booster samples (n=808) of ethnic groups. Of all respondents, 15 percent reported committing a property offense in the last year; 13 percent reported committing a violent offense; and 11 percent stated they had committed a handling offense (fraud or using stolen goods). These figures varied by age. Just under one-sixth of 16 and 17-year-olds and just over one-fifth of 18- to 22-year-olds reported committing a property offense in the last year, compared to only 7 percent of 23- to 25-year-olds; 16 percent of 18- to 21-year-olds said they had committed a handling offense in the last year; only 5 percent of 22- to 25-year-olds reported they had committed the same kind of crimes; 17 percent of 16- to 17-year-olds reported committing a violent offense in the last year, and just over 5 percent of 23- to 25-year-olds said they had committed similar offenses. At all ages, the crime-age profiles for the less educated (those who left school at age 16) differed from the profile for the more educated group (those who stayed in school past the compulsory age). The study then considered other variables that might relate to crime and age, i.e., neighborhood/area, school, individual, family, and labor market. The crime-age profiles for the less educated showed little sign of decline from around the age of 22. By age 25, this group had a much higher probability of committing all three types of offenses than the more educated group. Once other variables were introduced into the equation, the gap between the two profiles was reduced. Thus, the study showed that age, although correlated with crime, also correlated with other variables such as schooling, individual, and family characteristics. 2 tables, 4 figures, 37 references, and appended definitions of variables and supplementary tables