Several theoretical models have been constructed to determine the effects of buisness simulation games (BSGs) on learning performance. Although these models agree on the concept of learning-cycle effect, no empirical evidence supports the claim that the use of learning cycle activities with BSGs produces an effect on incremental gains in knowledge and skills. Therefore, this study aims to provide quantitative evidence by conducting an experiment using BSGs with different complexity levels in an undergraduate general course. Three research propositions guide a corresponding experimental design to collect data on student perception. Statistical analyses of the perception data from 43 student respondents reveal that skills, declared knowledge, prodedural knowledge, and strategic knowledge greatly increase in the initial cycles but vary in the rate of increase at later cycles. This research, which adds to the literature of game learning models, concludes that learnings cycles with different activities are required to stustain the development of knowledge and skills acquired through BSGs. The empirical outcome also serves as a useful reference for teachers who are planning to adopt BSGs in their class activities.