It is proffered rather frequently that co-offending is the dominate form of criminal offending among juveniles because of the enhanced salience of peer pressure during adolescence, and that this enhanced propensity to co-offend is pivotal for understanding the age-crime curve. Using National Incident-Based Reporting System (NIBRS) data for 2002, the authors conduct an analysis of 466,311 criminal arrests drawn from seven states. Their findings indicate that co-offending patterns by age are not noteworthy in elucidating why participation in illegal activities rises in adolescence, peaks in early adulthood, and then declines thereafter. Once co-offending is differentiated from solo offending, with solo offending representing the bulk of criminal activity among all age groups, including juveniles, a curvilinear relationship remains between age and solo-offending and between age and co-offending. These nonlinear associations are not conditioned by an offender's sex, race, or by offense type. The authors also analyzed co-offending crimes reported to police. In many types of crimes, offenders and victims come into contact, thus allowing for the estimation of the perpetrator's age notwithstanding whether an arrest was made. The findings generated in this supplemental analysis are similar to those produced using the arrest data. The results of this study have consequential theoretical implications not only because they cast doubt on the assertion that differences in co-offending levels between juveniles and adults account for the age-crime curve, but they also contravene the widely held belief that most juvenile offenses are perpetrated in the company of others rather than by individuals acting alone. Based on the new data reported here it appears that group offending is merely incidental in circumstance and thus of little etiological significance. (Contains 9 figures and 9 notes.)