In addition to serving summative assessment purposes, testing has turned out to be a powerful learning tool. However, while the beneficial effect of testing on learning performances has been confirmed in a large body of literature, the question of exactly how testing influences cognitive and metacognitive processes remains unclear. We therefore set out to investigate the effect of testing on self-regulated learning (SRL) processes. We hypothesized that by recalibrating metacognitive monitoring, regular practice testing can trigger efficient SRL processes and, in turn, foster learning. To test this hypothesis, we exposed first-year undergraduates to a complex neurology module. Participants were randomly assigned to either the practice testing group or a control group. The testing group underwent multiple practice tests during the neurology module, whereas the control group only underwent the multiple practice tests after the course. To assess the impact of practice testing on SRL processes, we combined a think aloud protocol with a metacognitive monitoring self-report measure. Results showed that, compared with controls, participants in the practice testing group were significantly less overconfident in their ability to recall recently learned information and performed better on a posttest questionnaire. Furthermore, mediation analyses confirmed that enhanced learning performance was explained by the use of efficient SRL processes. Therefore, these results allow us to extend the testing effect to SRL, and empirically underscore the central role of monitoring in SRL. Contributions to the fields of practice testing and SRL are discussed.