This article presents findings from a multisite cluster randomized trial of a structured organizational change intervention for improving HIV testing services in jails and prisons. Matched pairs of prison and jail facilities were randomized to experimental and control conditions; all facilities received baseline training about best practices in HIV testing and other HIV services and selected an area of HIV services on which to focus improvement efforts. The experimental facilities formed local change teams and were provided external coaching based on the Network for the Improvement of Addiction Treatment (NIATx) process improvement model. Difference-indifference analyses indicate a significant relative increase in HIV testing in the experimental compared to the control condition. Meta-analyses across the matched pairs indicated a small to medium effect of increased testing overall. The results indicate that the local change team model can achieve significant increases in HIV testing in correctional facilities. Implications for HIV testing policies and challenges for expanding testing are discussed.