Eastridge et al. (2012) reviewed US Service Member deaths between 2001 and 2011 and found that nearly a quarter of these deaths were potentially survivable, and the majority of those were largely attributable to missed and incorrectly performed interventions for hemorrhage injuries and airway obstruction. The focus of this paper is to describe an approach to improve the empirical assessment and therefore design of Human Patient Simulation (HPS) for training lifesaving airway and hemorrhage interventions. Every action taken by a combat medic has a corresponding outcome for the patient and is in response to the medic's continual assessment of that patient. There are identifiable cues (visual, auditory, or tactile) for successful and unsuccessful airway and hemorrhage interventions, and medics must be capable of detecting them. As more training is performed using HPS devices, the ability to support both procedural practice and patient assessment skills at the right level of fidelity become increasingly important. Systematic methods are necessary for identifying and addressing deficiencies in HPS and their corresponding impact on training effectiveness. Four considerations relevant to the desired specificity of data are included in the proposed systematic approach to HPS assessment: mapping fidelity assessment to training objectives, representing patient assessment cues, representing variations in injuries and pathologies, and converging measures of trainee performance. While this approach emphasizes a systematic approach to HPS assessment for the training audience of interest (combat medics), the aim is that these considerations will be broadly applicable for medical training.