Gaming-simulations (G-S) are those in which participants play roles, make decisions and receive feedback while interacting with a simulation model. This study used attitude change to measure of G-S effectiveness in exploring questions relating simulation design and effectiveness. To define participant attitudes and the content for a simulation, the organizational culture of a software engineering firm was studied using an ethnographic approach. Inconsistencies between the existing culture and expressed ideals were measured using an 40 item attitude questionnaire drawn from statements made during interviews. Simulation structure and participant cognitive style were factors hypothesized to influence identification with a simulation role. Role identification was hypothesized to influence attitude change. Two versions of the simulation were designed to produce differential role identification. Role identification was measured by having the simulation software ask players questions near the end of the eight hour simulation class. The Davis (1980) Empathy scale was used to measure cognitive style. The WINNING AT DESIGN AUTOMATION gaming-simulation was created to induce attitude change toward the ideals. The simulation is written in HyperCard. Each participant managed a department, allocating their time to tasks that earned points as they competed in teams. A control group of 42 employees and 97 of the 122 who played the simulation completed pre- and post-simulation questionnaires. There was significant attitude change for all treatment groups. The different versions of the simulation did generate stronger and weaker role identification as predicted. The Empathy scale did predict role identification. The treatment group with highest role identification did not have the greatest attitude change. Other factors influencing the linkage of role identification to attitude change are discussed.