The importance of pharmacokinetic and receptor studies in the preclinical and clinical safety evaluation of candidate drugs is reviewed with reference to a number of recently developed drugs. Different aspects of the relationships between pathways of metabolism, pharmacokinetics, receptor interactions, and drug toxicity are illustrated. The failure of animal toxicity studies to predict drug toxicity in humans, due to species differences in metabolism and pharmacokinetics, is illustrated by reference to the anti-inflammatory antiviral terpenoid carbenoxolone, the antiasthmatic candidate drug FPL 52757, and the cardiotonic drug amrinone. The false prediction of adverse effects in man from toxicity manifested in experimental animals, due to species differences in pharmacokinetics or receptor activities, is exemplified with reference to the antiepileptic valproic acid, the hypolipidemic drug ciprofibrate, the antipeptic ulcer durg, omeprazole, and the progestogen lynestrenol. Finally, the importance of adequate, repeat-dose, clinical pharmacokinetic studies in patients as distinct from healthy volunteers to evaluate any effect of the disease state, in the elderly and the young to examine the effects of age, and in sufficiently large populations to detect genetic anomalies and idiosyncrasies is illustrated by reference to the antirheumatoid drug benoxaprofen, the antiangina drug perhexiline, and the diuretic tienilic acid.