Artificial intelligence (AI) systems are increasingly embedded into daily life, shaping decisions in domains ranging from hiring and education to healthcare and criminal justice. While these systems promise efficiency and scale, they also risk embedding and amplifying existing social inequalities. In response, AI auditing has emerged as a key accountability mechanism, uncovering harmful algorithmic behaviors such as bias, discrimination, and misrepresentation. Traditional audits, however, have been largely expert-driven, conducted either by in-house teams or external researchers. Although impactful, these audits often fail to capture harms that only end users—those directly affected by AI systems—can recognize in situated, real-world contexts. This chapter introduces and develops the concept of participatory AI auditing (PAIA): an approach that meaningfully engages end users in the auditing process to surface harms, contextualize findings, and foster accountability. This chapter begins by situating PAIA within the broader lineage of AI auditing and participatory design, then synthesizes current challenges in enabling effective PAIA. The chapter further surveys emerging tools and processes designed to support PAIA, and closes by envisioning incorporating PAIA as a central piece to human-centered AI, where systems are designed not only for but also with the people most affected by them.