Abstract: The parton distribution functions (PDFs) of the proton are normally fitted to data under the assumptions of the Standard Model (SM); in particular, subsequently using these SM PDFs in a search for beyond the Standard Model (BSM) physics could produce misleading results. Instead, the proper course of action should be to simultaneously determine PDFs with parameters in the BSM model under study; in some cases, the difference between the resulting BSM PDFs and SM PDFs can be large. I will discuss these issues in the context of the joint determination of PDFs and parameters in the Standard Model Effective Field Theory (SMEFT), especially relating to high-mass Drell-Yan data and top data, and (if time permits) in the context of a light dark matter model.