Studies suggest that saliva could be used instead of blood in the therapeutic monitoring of many drugs. This has distinct advantages in pediatrics and neonatology as saliva sampling is painless and spares blood. Stimulation of saliva secretion with a chemical stimulus (i.e. citric acid applied over the tongue) facilitates the study of younger patients. Secretory and reabsorptive processes which take place in the ductal system of the salivary glands, and the rate of flow of the secretion play major roles in the determination of the concentration of solutes in saliva. Drug passage into saliva follows the general principles of movement of drugs across biologic membranes. Only the unbound fraction of the drug in plasma is available for diffusion into saliva and a relationship exists between saliva pH and the saliva/plasma concentration ratio of many polar drugs (tolbutamide, propranolol, procainamide, etc.). However, deviations from the pH theory exist and the inter -and intra-individual variations in saliva/plasma concentration ratios of salicylate and procainamide cannot be explained solely on the basis of fluctuations of salivary pH; on the other hand, a useful relationship exists between plasma and saliva phenobarbital concentrations with no need to correct for saliva pH. The use of stimulated saliva has several advantages over resting saliva: a larger volume of the sample is obtained, the pH gradient between plasma and saliva is smaller, the variability in saliva/plasma concentration ratios of some drugs is narrowed, and less specimens are too viscous or discolored to allow drug analysis. Thorough rinsing of the mouth is required prior to saliva sampling as remnants of orally administered medicines may contaminate saliva specimens and give spuriously high values. Deviation from a simple but strict methodology accounts for some of the discrepancies found in the literature. Studies in children uniformly recommend saliva for therapeutic monitoring of phenytoin, carbamazepine and phenobarbital. Saliva sampling for therapeutic monitoring of ethosuximide, primidone and digoxin in infants and children, and of theophylline and caffeine in the neonate is promising, but little pediatric experience is available as yet. The value of saliva in therapeutic monitoring of theophylline in children is still controversial. Little of highly polar compounds such as aminoglycosides, and of polar highly protein bound drugs such as valproic acid is present in saliva. More data are still needed on the excretion of drugs in saliva in infants and in acutely ill children, and few data exist in the premature and full-term neonate.