There is considerable research asserting the importance of exploratory search versus exploitative search, however relatively little work examines strategies used to achieve exploratory search. How do individuals and firms explore uncharted technological terrain? This paper extends research on knowledge networks, cognition, and learning to propose three main strategies of knowledge creation that are more likely to result in discoveries that are distant from existing inventions: Deduction, synthetic recombination, and long search paths. We test our arguments using a large and unique dataset on outlier patents filed at the US Patent and Trademark Office. Our findings suggest that there are significant differences in the inventor teams, assignees, and search strategies that result in outlier patents. These results have important implications for managers that wish to encourage more exploratory search for breakthrough innovation.