The predictive efficacy of drugs in humans is frequently estimated from both a high affinity for their receptor as measured in vitro and a long plasmatic half-life. This is grossly misleading since one key parameter is missing: drug concentration at the receptor site in vivo. As a case study we compared the efficacies of three H(1) antihistamines in inhibiting histamine-induced wheal and flare in humans at two different time points with the above mentioned parameters. It is concluded that estimating in vivo receptor occupancy, which takes into account both the affinity of the drug for the receptor and its free plasma concentration, is a far better predictor for human pharmacodynamics and hence antihistamine potency, than considering in vitro affinity and plasmatic half-life only.