The absence of selective antagonists makes receptor characterization difficult, and largely dependent on the use of agonists. However, there has been considerable debate as to whether certain drugs acting at G protein-coupled receptors are better described as agonists, partial agonists or antagonists. Indeed many agonists tend to display differences in intrinsic activity, depending on the preparation used to study receptor pharmacology. It has been argued that variations in intrinsic activity of drugs, may be a reflection of receptor subtypes rather than varying degrees of receptor-effector coupling. However, it has often been overlooked that classical receptor theory predicts that a drug acting at a given receptor type can display a range of intrinsic activities depending on the extent of receptor reserve. Furthermore, because any receptor can couple to different effector systems, different degrees of intrinsic activity may be observed for a given drug. Daniel Hoyer and Hendrikus Boddeke review recent data which provide explanations for why drugs display such variations in intrinsic activity.