One of the underlying tenets of clinical pharmacology is that only free drugs are pharmacologically active. It is thought that only free drugs can cross biological membranes to interact with a given receptor to alter its function, and that drug responses, both efficacious and toxic, are a function of unbound concentrations. The rationale for measuring drugs in oral fluid is that the free fraction of a drug in plasma reaches equilibrium with the drug in saliva. Although reports concerning the appearance of organic solutes in saliva have been in the literature for over 70 years, it has only been in the past 30 years that there has been emphasis on the appearance of drugs. Although many assumptions for drug level monitoring in saliva are made, the primary requisite for salivary monitoring to be useful is a constant or predictable relationship between the drug concentration in saliva and the drug concentration in plasma. Measurement of oral fluid drug levels for the purpose of managing patients and making dosage adjustments may be useful for select drugs or drug classes. However, it does not appear to be useful for the majority of drugs therapeutically monitored. Some work with antipsychotic medications has indicated that although the measurement of drug concentrations themselves may not be useful for dosage adjustment, the ratio of parent drug to metabolite may reflect altered metabolic status due to either pharmacogenetic variation or other clinical conditions. Furthermore, analysis of saliva may provide a cost-effective approach for the screening of large populations.