This survey discusses the correlation between the oral potency of antialgesic drugs in several pharmacology laboratories and their human oral dose in clinical practice. We also present a brief overview of a few biological assays that have been successfully used to direct the synthesis of newer antialgesic drugs. The laboratory assay that our analysis showed to be most predictive of the clinical analgesic dose is based upon the response of rats to flexion of an arthritic joint. Laboratory ED50 values from the ACh-induced abdominal constriction assay in mice are nearly as predictive while the predictive power of the yeast-induced hyperalgesia assay in rats is somewhat less. Probably because of the small number of experiments, the correlation between the efficacy of these agents in a canine model of synovitis and their clinical doses only reached borderline statistical significance (p = 0.0651). Regression equations are presented that permit calculations of single clinical analgesic doses from efficacy data in individual tests. Calculation of stepwise multiple regression showed that the clinical dose could be best predicted when efficacy data obtained in the joint flexion assay in rats and the ACh-induced constriction assay in mice are both taken into account. We have concluded that the effective doses are highly predictive of clinical efficacy because these animal assays have been designed to reflect the action of drugs upon prostanoid-induced hyperalgesia.