Criminal justice researchers have studied caseload size to determine whether smaller caseloads improve probation outcomes. The purpose of this study was to investigate whether reduced caseloads improved supervision outcomes for medium to high risk offenders in a probation agency that trains its officers to apply a balance of controlling and correctional/rehabilitative measures. Three different probation agencies were selected to take part in this study. The first was Oklahoma City (Datasets 1 and 2), where a randomized controlled trial (RCT) experiment was implemented. The second site was Polk County, Iowa (Dataset 3), where a regression discontinuity design study (RDD) was implemented. Lastly, four judicial districts in Colorado (Dataset 4) were selected where a RDD study again was implemented. In Oklahoma City the RCT degenerated and the study team turned to a difference in differences (DD) estimator. The research team also collected Evidence-Based Practices Probation Officer Data (Dataset 5) from a set of survey questionnaires administered to the participating officers from the three research sites and the scoring of taped officers' supervision sessions with probationers on their caseloads by trained raters.