Delivery of services and programs to individuals who are incarcerated and those who have been released recently often falls short of expectations formed during the program planning stage. When programs are partially implemented, program participants receive less than what was prescribedfewer services and lower dosages of provided services. Partial implementation has the effect of diluting treatment effects and reducing the power of outcome evaluations under an intent-to-treat analysis. This article uses data collected from surveys of 12 reentry program directors and interviews with 863 adult male prisoners participating in these 12 individually designed, comprehensive reentry programs. The programs were implemented using Federal funds obtained through Serious and Violent Offender Reentry Initiative (SVORI) grants and were designed to provide a variety of services. This article describes the prelease services program directors identified as components of their local SVORI programs and compares these site-level reports with the services participants reported receiving. The analyses provide a "first-order" measure of implementation that assesses the extent to which those providing services and those receiving services agreed in terms of the prevalence of service receipt. The hypothesis is that agreement suggests that the program was at minimum delivering the services planned. The authors then look in more detail at the variation with respect to the delivery of 21 specific prerelease service items, individually and as part of constructed "service bundles." Results suggest considerable variability across the sites and across service items and bundle scores in both levels of service reported and the extent of agreement between program directors and participants. (Published Abstract)