Most of the knowledge generated to bridge the research-practice gap has been derived from experimental studies implementing specific treatment models. Alternatively, this study uses observational methods to generate knowledge about community-based treatment processes and outcomes. Aims are to (a) describe outcome trajectories for children with disruptive behavior problems (DBPs), and (b) test how observed delivery of a benchmark set of practice elements common in evidence-based treatments may be associated with outcome change while accounting for potential confounding variables. Participants included 190 children ages 4 to 13 with DBPs and their caregivers, plus 85 psychotherapists, recruited from six clinics. All treatment sessions were videotaped and a random sample of 4 sessions in the first 4 months of treatment was reliably coded for intensity on 27 practice elements (benchmark set and others). Three outcomes (child symptom severity, parent discipline, and family functioning) were assessed by parent report at intake, 4, and 8 months. Data were collected on several potential covariates including child, parent, therapist, and service use characteristics. Multilevel modeling was used to assess relationships between observed practice and outcome slopes while accounting for covariates. Children and families demonstrated improvements in all 3 outcomes, but few significant associations between treatment processes and outcome change were identified. Families receiving greater intensity on the benchmark practice elements did demonstrate greater improvement in the parental discipline outcome. Observed changes in outcomes for families in community care were generally not strongly associated with the type or amount of treatment received.