This research sought to determine whether learners would increase their knowledge of six types of accounting transactions using an interactive model of the accounting equation in the Multi-User Virtual Environment (MUVE) called Second Life. The experiment was conducted within the theoretical framework of both cognitive constructivism and social constructivism. Participants worked in one of two experimental treatment types—working solo (cognitive constructivism) or working in small groups (social constructivism). All participants used the same written instructions, instructional movies, and practice problems to learn about six types of accounting transactions. Practice problems were used in conjunction with a 3D model of the accounting equation in Second Life. The model provides immediate visual and text feedback to participants as they indicate their answers to practice problems. Participants can adjust their answers until the 3D model confirms the answer is correct. Results of pre- and posttests measured changes in participants’ level of knowledge of these six transaction types and changes in the level of participant anxiety during the experiment. To obtain two sets of data, all the steps of the experiment were completed twice. Each instance is referred to as a round of the experiment. The hypotheses predicted that, consistent with prior research in constructivism, participants who worked in small groups would experience greater learning gains and anxiety reductions than participants who worked alone. However, in both rounds of the experiment, participants who worked alone found both greater learning gains and greater anxiety reductions than those who worked in groups. This research resulted in a positive implication that Second Life can be used as a learning environment with reasonable expectation of learning. It also resulted in a new question: whether constructivism, as it is understood today, applies accurately to Multi-User Virtual Environments. Repetition of constructivist-based research in these environments is required to determine whether this result is consistent across populations and platforms. Expansion or revision of our understanding of constructivism may be required based on those findings.