A modified delphi group forecasting technique was used. A group of 13 subject-matter experts, stakeholders, and academics with varying backgrounds were enlisted for the group. Panel members initially selected and ranked several possible/probable future trends and events that would impact the central issue of this study over the next 10 years (1995 to 2005). The 10 most critical trends are outlined in this report, along with a listing of the top 10 events selected by the panel that could possibly or probably occur over the designated time period. The probability of occurrence of each event over this time span was then forecast, as was the probable impact of the event on the central issue of this study. To determine the final probability of occurrence and impact of each event, a cross-impact analysis was completed with the use of a computer program. The cross-impact data were then installed into another computer program that generated "scenarios." Three representative scenarios (best case, worst case, and exploratory) generated by the computer program were then selected to be expanded in narrative form. In addition, each of the 10 "trends" were assumed to have occurred as projected for the purposes of constructing the written version of each selected scenario. A strategic plan was developed for the implementation of selected policies that will result in the achievement of the goal in the management of the HIV/AIDS pandemic in the jail system. The strategy selected for further development and analysis was to establish early release programs or alternatives to incarceration within the jail system for HIV/AIDS infected inmates. This policy alternative appeared to have the potential for the broadest stakeholder support, coupled with the greatest potential for cost savings. Further, implementation of this policy would decrease the risk of infection to staff and inmates and encourage the testing and disclosure of HIV infection. An implementation plan and transition management for the strategy are discussed. 23 notes