Purpose: Health care expenditure on diagnostic imaging investigations is increasing, and many tests are ordered inappropriately. Validated clinical decision rules (CDRs) for certain conditions are available to aid in assessing the need for imaging. However, awareness and utilization of CDRs are lacking. This study compared the efficacy and perceived impact of interactive e-learning modules versus static versions of CDRs, for learning about appropriate imaging referrals. Methods: A multicenter, randomized, crossover trial was performed; participants were volunteer medical students and recent graduates. In week 1, group 1 received an e-learning module on appropriate imaging referrals for pulmonary embolism; group 2 received PDF versions of relevant CDRs, and an online quiz with feedback. In week 2, the groups crossed over, focusing on imaging referrals for cervical spine trauma in adults. Online assessments were administered to both groups at the end of each week, and participants completed an online questionnaire at the end of the trial. Results: Group 1 (e-learning module) performed significantly better on the pulmonary embolism knowledge assessment. After the crossover, participants in group 2 (e-learning module) were significantly more likely to improve their scores in the assessment of cervical spine trauma knowledge. Both groups gave positive evaluations of the e-learning modules. Conclusions: Interactive e-learning was significantly more effective for learning in this cohort, compared with static CDRs. We believe that the authentic clinical scenarios, feedback, and integration provided by the e-learning modules contributed to their impact. This study has implications for implementation of e-learning tools to facilitate appropriate referrals for imaging investigations in clinical practice. ? 2015 American College of Radiology.