Abstract More nontraditional, adult learners are entering higher education than ever before due to the continued growth in technology and current economic conditions that encourage adults to acquire additional skills for the workplace. The pressure for these students to do well in school and finish quickly can lead to increased anxiety associated with the high-stakes traditional exams used by many college instructors. This anxiety is a problem because for some adult learners, inflated stress levels negatively affect performance on exams and overall grades, therefore increasing the chances of dropping out. This study explored the potential relationships between adult learners' anxiety levels and their test scores from each phase of a 2-stage cooperative exam, an alternative to the traditional, individual high-stakes exam. The theoretical framework for this study included adult learning and constructivism. This nonexperimental and correlational research study used nonprobability, single-stage sampling. Using the Pearson r for correlations, results from the S-scale portion of the State-Trait Anxiety Inventory (STAI) were correlated with each phase of the students' cooperative scores. Results showed a significant inverse correlation (r = -.387) between adult learners' anxiety levels and their scores on the individual portion of a cooperative exam at a .05 alpha level, while showing no significant relationship between adult learners' anxiety levels and their scores on the group portion of the exam. This study supports using cooperative testing, thereby lowering adult learners' anxiety levels and improving the chances for achievement and graduation.