Regression analysis, using Japanese data, is used to test the hypothesis that crime rates depend upon returns to crime relative to returns to legal employment and the deterrent effects of the certainty and severity of punishment. The empirical results, while showing that the economic model is no more accurate than alternative naive models, do indicate that increases in returns to legitimate work diminish both adult and juvenile crime. Unemployment affects some adult crimes, but has little impact on juvenile crime rates. The proportion of the population living in poverty has no significant impact on either adult or juvenile crime. Finally, while increases in the certainty of punishment deters adult crime, there is but weak evidence that increasing the severity of adult punishment also deters. There is also little evidence that increases in the certainty of punishment deter juvenile crime. 7 tables, 35 references, and 1 appendix (Author abstract modified)