The goal for data-quality improvement should be the consistent achievement of acceptable levels of data errors, which experts in the field consider to be no more than 1 or 2 percent of the total. After providing an overview of the design of the outcome measurement system for Federal pretrial services, this paper summarizes what is being done to reach this goal. In 2005, the Office of Probation and Pretrial Services established a committee of chiefs, supervisors, officers, technical personnel, and data-quality analysts from probation and pretrial services offices in various districts. This working group is focusing on how data quality should be defined and how to communicate this information to the districts. It has established a Web site that provides standard data-quality reports to the districts in an effort to achieve compliance with a national standard. The strategy is for each district to create its own data-quality improvement program. In advancing this effort, the data-quality working group developed the District Data Quality Program Development Guide, which provides a step-by-step process for a district in developing its data-quality program. This paper outlines the suggested steps. Planning for training in data-quality improvement has focused on how to prepare data-entry staff to enter data accurately and how to identify data entered inaccurately. In an effort to monitor data-quality improvement in the districts, the Office of Probation and Pretrial Services attempts to conduct 20 program reviews annually. These reviews are designed to assist districts in identifying and addressing problems in existing policies and practices for collecting and entering data. 1 figure