Behavioral parent training is an effective treatment for many child behavior problems; however, low parent attendance and engagement has been a chronic barrier to its successful implementation. Previous research identified a number of risk factors that were predictive of low engagement in parent training. The present study tested whether these risk factors were valid predictors in a targeted prevention sample using latent class analysis and a binary segmentation procedure to identify meaningful sub-groups within the sample. Although the latent class analysis did not identify meaningful classes which predicted attendance, the binary segmentation procedure resulted in six mutually exclusive groups. These groups were classified based on social support and stressful life events, and group membership significantly predicted attendance at parent training. The validity of these predictors was further supported by a backward stepwise regression. Other frequently studied predictors, such as income, did not discriminate within the sample. These findings suggest that the risk factors for low engagement and participation in targeted prevention parenting interventions may be different from the risk factors for treatment seeking samples.