The credibility of US Army analytical experiments using distributed simulation depends on the quality of the simulation, the pedigree of the input data, and the appropriateness of the simulation system to the problem. The second of these factors is best met by using classified performance data from the Army Materiel Systems Analysis Activity (AMSAA) for essential battlefield behaviors, like sensors, weapon fire, and damage assessment. Until recently, using classified data has been a time-consuming and expensive endeavor: it requires significant technical expertise to load, and it is difficult to verify that it works correctly. Fortunately, new capabilities, tools, and processes are available that greatly reduce these costs. This paper will discuss these developments, a new method to verify that all of the components are configured and operate properly, and the application to recent Army Capabilities Integration Center (ARCIC) experiments. Three recent developments have focused improving the process to load the data. OneSAF has redesigned their input data file formats and structures so that they correspond exactly with the Standard File Format (SFF) defined by AMSAA, ARCIC developed a library of supporting configurations that correlate directly to the AMSAA nomenclature, and the Entity Validation Tool was designed to quickly execute the essential models with a test-jig approach to identify problems with the loaded data. The missing part of the process is provided by the new Expected Results Method. Instead of the usual subjective assessment of quality, e.g., “It looks about right to meâ€�, this new approach compares the performance of a combat model with authoritative expectations to quickly verify that the model, data, and simulation are all working correctly. Integrated together, these developments now make it possible to use AMSAA classified performance data with minimal time and maximum assurance that the experiment's analytical results will be of the highest quality possible.