Despite the overwhelming evidence for dark matter in our Universe, very little is known about its properties. A very well motivated hypothesis is that the dark matter is constituted by long-lived massive particles which interact very weakly with the Standard Model particles. If this is the case, the dark matter could interact with the nuclei of a dedicated detector, in a search strategy dubbed "direct detection". From the theory side, direct detection experiments are plagued with uncertainties from our ignorance of the dark matter velocity distribution inside the Solar System and from our ignorance of the concrete from of the dark matter interactions with the nuclei. In this talk, I will review these uncertainties and I will present some recent new ideas to bracket theoretical uncertainties when interpreting the experimental results.