Research presented in this article addresses some of the competing cla
ims about the value of prospective longitudinal self-report data for s
tudying the relationship between age and crime More specifically, the
debate over the magnitude of potential testing effects in longitudinal
data is assessed by analyzing involvement in delinquency, serious off
ending, and victimization using self-report data from the first five w
aves of the National Youth Survey. The results of growth curve analyse
s suggest that panel and maturation effects warrant serious concern in
longitudinal studies that rely on self-report information. The analys
es show that regardless of the subject's age at the start of data coll
ection, average self-reported involvement in crime declined substantia
lly over time. Data from external sources suggest that period effects
are not responsible for this decline. Furthermore, while traditional m
easures of delinquency are reliable for studying between-individual di
fferences in crime these same measures may lack the reliability necess
ary for studying between-individual differences in changes in crime ov
er time.