Attrition is a significant problem in child psychotherapy (De Haan, Boon, De Jong, Hoeve, & Vermeiren, 2013) and has serious implications. These include limited effectiveness of interventions for mental and behavioral health problems and decreases in staff productivity, access to services, and the number of people an agency can serve (Barrett et al., 2008). Predictors of attrition include those at the client, family, therapist, and treatment levels. The current study investigated the pattern of and variables associated with attrition in Parent-Child Interaction Therapy (PCIT), an evidence-based parent training program for families with young children (ages 2.5-to 7-years-old) with disruptive behavior disorders. Client-(i.e., child gender, child minority status, intensity of externalizing behavior); parent-(i.e., parent income, parent-to-child ratio, depression symptoms); and treatment-level (wait time for treatment, parent satisfaction with PCIT, therapist satisfaction with PCIT) variables; and competing activities and life stressors were examined. The sample of 134 parent-child dyads and 36 therapists recruited from 21 licensed psychiatric outpatient clinics across one state were part of a larger study (RO1 MH09750; A Statewide Trial to Compare Three Training Models for Implementing an EBT; PI: Herschell). The current study found an attrition rate of 73%. Families were more likely to leave early in treatment, with 53% of families leaving therapy after attending six or fewer sessions. Hierarchical Linear Modeling was used to examine predictors. Lower income, a smaller ratio of parents to children, longer wait time for PCIT, more reported life stressors and obstacles, and lower parent satisfaction with PCIT were all significantly associated with attrition. Qualitative data regarding additional therapist-reported reasons for attrition are reported. Limitations, suggestions for future research, and clinical implications are discussed.