Transfer of learning is a fundamental goal of education but is challenging to achieve, especially where far transfer to remote contexts is at stake. How can we improve learners' flexible application of knowledge to distant domains? In a counterintuitive phenomenon termed the "derring effect," deliberately committing and correcting errors in low-stakes contexts enhances learning more than avoiding errors. Whereas this benefit has been demonstrated with tests in domains similar to those in the initial learning task, the present set of three experiments (N = 120) investigated whether deliberate erring boosts far transfer of conceptual knowledge to dissimilar domains. Undergraduates studied scientific expository texts either by generating conceptually correct responses or by deliberately generating conceptually erroneous responses then correcting them. Deliberate erring improved not only retention (Experiment 1), but also far transfer on inferential test questions that required applying the learned concepts to remote knowledge domains (e.g., from biology/vaccines to geography/forest management techniques; Experiment 2). This advantage held even over a control that further involved spotting and correcting the same errors that one's peers had deliberately made (Experiment 3). Yet, learners failed to predict or recognize the benefits of deliberate erring even after the test. Altogether, these results suggest that the derring effect is specific to generating incorrect, but not correct, elaborations. Neither does mere exposure to others' errors nor juxtaposing these errors with the correct responses suffice. Rather, guiding learners to personally commit and correct deliberate errors is vital for enhancing generalization and far transfer of learning to distant knowledge domains.