This research is funded by the Lahore School Research Development Fund (LSRDF). Despite the introduction of several anti-corruption interventions in recent years, corruption remains ubiquitous in developing countries. The likelihood of successful anti-corruption reform would involve answers to two basic questions. The first and, may be the most basic question is: how prevalent is corruption? The second question is: what are the effective approaches that can be brought to limit corruption in public offices? This project attempts to shed new light on these two questions by using unique survey experiments with Pakistani bureaucrats. We use sensitive survey techniques, which take into account the social desirability bias associated with corruption. More precisely, we use two list experiments to estimate the rate of overall corruption and misuse of public funds in the Pakistani bureaucracy, and we use these list experiments and an endorsement experiment to test the effectiveness of four anticorruption policy interventions—wage incentives, community monitoring, auditing, and social sanctions.