Herein, we describe a approach to retrieve free tropospheric columns of peroxyacyl nitrates (PANs) from radiances observed by the Atmospheric Infrared Sounder (AIRS). AIRS has provided daily global coverage since its launch in 2002, making the AIRS data a valuable long term record. Although the instrument is very radiometrically stable, the radiance noise level is large enough to present a challenge when retrieving a weak absorber such as PAN. To address this, we focus on retrievals over land (to avoid interferences from low, warm clouds over ocean) and develop a decision tree quality filter trained to predict whether a PAN value retrieved from AIRS will be within 0.2 ppb or 50 % of what would be retrieved from the Cross-track Infrared Sounder (CrIS). We show that AIRS is capable of retrieving PAN plumes from significant wildfires that match those retrieved from CrIS and that PAN retrieved from AIRS has good correlation with CrIS given sufficient averaging. We conclude with recommendations for users to help ensure that these data are used appropriately.