Computer-supported training simulations have been recognized for the potential and the benefits they have in supplementing the training needs of the military, yet we still do not see evidence of large-scale deployment and adoption of these systems by users in this domain. The current challenging budgetary situation suggests that the Return on Investment (ROI) will be more scrutinized than ever before, forcing communities to abandon underutilized and underperformed Modeling and Simulation (M&S) solutions. Such developments are also likely to affect global decisions related to future investments in these types of technologies. This paper presents the design and results of a study that included collection of comprehensive data on the adoption and use of training simulations in the military domain. The analysis of this data set suggests that the reasons for low use of simulations had little to do with the overall quality of hardware and software (although they were mentioned as factors), and that a myriad of other factors were found to influence the outcome to a greater extent. The understandings collected in this and other studies all attest that military training is a complex, multilayered domain that is only partially defined by the type and technical characteristics of systems being used to achieve that goal. Our work and experience in this domain give us a firm basis to hypothesize that a well selected set of strategic approaches could bring much greater results in this domain, even with a modest investment made to support that change. The findings and recommendations are highly applicable to all DoD services and other communities that plan to use these types or solutions in their training and learning practices. The study also offers a contribution towards a better understanding of general diffusion and adoption of other technical innovations in the military domain.