Abstract: "The LHC is a hadron collider, and the detailed understanding of strong-interaction dynamics is thus an important ingredient for its successful physics programme. In view of the increasingly involved strategies relevant to the majority of BSM searches, the command of accurate predictions helps reduce the theoretical bias on experimental analyses in a significant manner. Several key methods have been established in QCD in the recent past, which have led to a general increase of precision in the computation of collider physics observables. I shall discuss some of the theoretical ideas which underpin such methods, illustrate selected applications to LHC physics, and briefly touch on implications for the near future."