Since the 1985 Emergency Cardiac Care Conference, numerous controversies about the pharmacology of CPR have arisen (eg, questions about the pharmacokinetics and pharmacodynamics of drugs during CPR, the optimal vehicle for delivery of medications, and the dose of atropine in brady-asystolic cardiac arrest). This article has three objectives: 1) to critically explore these controversies, 2) to provide recommendations for clinical practice, and 3) to identify areas for future study. The ideal route is one which combines rapid access with quick delivery of drug to the central circulation. Because of hemodynamic changes during CPR, administration of drugs into the central circulation is preferable when compared with peripheral venous injection. Whenever drugs are administered from a peripheral i.v. site, the extremity should be elevated, and a 20-mL bolus of i.v. fluid should be given to facilitate access of the agent to the central circulation. If there is a delay in obtaining venous access, epinephrine, lidocaine, and atropine may be administered through the endotracheal tube at 2.5 times the i.v. dose. When administering these drugs through the endotracheal tube, dilute the drug in 10 mL of saline or water and inject it through a long catheter beyond the tip of the endotracheal tube. Dextrose 5% water is the primary vehicle for drug delivery during CPR. However, the administration of glucose during CPR is controversial because of the potentially detrimental effects of hyperglycemia on neuronal function during periods of ischemia. Data are inconclusive regarding the effects of glucose levels on neurologic outcome following resuscitation. Hyperglycemia may be a marker for prolonged resuscitation with subsequent impairment in insulin release.(ABSTRACT TRUNCATED AT 250 WORDS)