Fields within education and training have been exploring the use of educational computer-based games, often referred to as serious games (SG), in multiple disciplines of academic research including the affective, cognitive, and psychomotor domains. Traditionally, game designers tend to represent a different viewpoint about learning than instructional designers, or even teachers. More so, one of the fundamental roles designers play in making decisions is based on multiple factors, which include personal assumptions about constraints and perceived constraints in instructional practice. In order for games to be successful in classroom environments, classroom teachers need to be involved in the design process to help identify and assist in mitigating the classroom-based challenges that will be faced during implementation. The study sought to extend research on serious game attributes by examining the instructional design decisions and beliefs of individuals involved in the design, development, or implementation of serious games in education or training environments, through a web-based survey. Within the serious game community there are multiple approaches to designing learning environments; some view serious games as virtual environments explicitly for education or training, while others include digital games, simulations, and virtual worlds. While there is debate over the type of games that are most effective for learning, researchers have provided guiding qualifications and lists of characteristics that effective games should possess to improve current practice and implementation. Two central aims guided the study: (a) to identify relationships between the mental models put forth by each discipline when selecting serious game attributes, and (b) to provide insight into each subpopulation's beliefs about learning. Suggested implications for the study extend to educational practice, policy, and future research on designing, developing, and implementing serious games in learning environments. Findings suggest that the sample portrayed similar epistemological beliefs between all subgroups. Participants had the most sophisticated beliefs toward quick learning. Limited relationships were evident between participant's epistemological beliefs and selection of serious game attributes (SGA). However, findings indicated that each discipline has unique models and frameworks for designing serious games and perspectives on serious game implementation.