This study investigated (a) the efficacy of using descriptive and experimental assessment methodologies to generate hypotheses regarding the function of problem behavior and (b) the efficiency and efficacy of function-based interventions compared to traditional intervention approaches that focus on the topography of behavior. Functional assessments were conducted with three elementary school students identified as at risk for failure due to problem behavior. Agreement among indirect measures, direct observation, and experimental manipulation of environmental variables supports the value of using convergent data from indirect assessment methods to develop valid hypotheses. In addition, behavioral interventions based on functional assessment were found to be more effective than alternative intervention approaches across all three case studies. Implications, study limitations, and future research directions are discussed.