Various studies have examined why people commit sexual offenses, but most have dealt with adult offenders. An increasing number of studies show that juveniles are committing sexual offenses at higher rates than in the past. Studies also suggest that many offenses continue to go unreported. As with other violent crime, the vast majority of juvenile sex offenders are male. I investigate juvenile sex offenders from the perspective of General Strain Theory, using data obtained from 253 juvenile males in state custody in Oklahoma during the summer of 2001 (the Oklahoma Juvenile Data). I compare two subsets of juvenile males, those who self-report some sort of sexual offense (molestation or rape) during their lifetime (n = 47) to those who self-report committing another type of violent offense (n = 193). According to General Strain Theory, there are three major sources of strain: the failure to achieve positively valued goals, the loss of positive stimuli, and the presentation of negative stimuli (Agnew 1985). I identify and examine sources of strain, as well as other factors, which may help explain the difference in violent but non-sex offending juveniles and juvenile sex offenders. The findings from this exploratory study will be helpful in developing programs for reducing sex offenses among youth.