Prior research finds that risk-taking has declined after the Sarbanes-Oxley Act of 2002, consistent with the notion that SOX's corporate governance and internal control mandates diverted resources away from corporate risk-taking. We introduce to the accounting literature a new measure of R&D productivity, Research Quotient, to examine whether SOX affects R&D risk-taking and R&D productivity differently and whether the quality of the firm's governance and internal controls, pre-SOX, moderate these relations. While we find the relation between SOX and R&D risk-taking is sensitive to research design choices, we find a consistent positive relation between SOX and Research Quotient. Our evidence indicates that while firms may allocate fewer resources to R&D post-SOX, they concurrently manage their R&D investments more productively. Further, our results are robust to a difference-in-difference design and are stronger for firms with weaker governance pre-SOX.