In 2009 and 2010, the Gujarat Pollution Control Board in India used a third-party audit system for plants with high potential to pollute as part of its regulatory program. Under this system, auditors visited the plant and took samples three times over the course of a year, then submitted an audit report to the Pollution Control Board that could serve as the basis for regulatory action. The audit system incorporated several safeguards by requiring auditor accreditation, limiting an auditor from accepting consultancy work from the plants they audited, limiting the number of audits undertaken in the year, and granting authority to decertify auditors found to be inaccurate. Despite these safeguards, an experiment designed to measure the effects of the auditors’ pay incentives revealed striking results. In the first year of the experiment, auditors were randomly assigned to a group of plants (the “treatment group”), paid through a central account, and informed that their audits could be subject to verification. In the second year, auditors assigned to the treatment group were informed their pay for an audit would be scaled based on its accuracy. The “control group” of auditors continued to be paid by plants directly and was not told that their audit could be subject to backchecks. The control group systematically underreported pollution readings, compared to the results as measured by backchecks. Notably, auditors in the control group systematically and incorrectly reported many pollution readings to be just below the regulatory standard (i.e., in compliance). In the treatment group, on the other hand, the changes in pay incentive structure resulted in the audits reporting results consistent with backchecks by the end of the experiment. More remarkably, the plants that were subject to increasingly accurate audit reports responded by significantly reducing their pollution emissions. Thus, performance-based pay incentives not only improved employee performance, they improved environmental outcomes.