Hexobarbital (HB) concentrations were determined in plasma and saliva of 8 healthy subjects, following oral administration of 500 mg HB-Na. Mean plasma half-lives were 3.2 +/- 0.1 h, and salivary half-lives 3.3 +/- 0.2 h. Mean plasma clearance was 22.9 +/- 2.3 1 h-1. There was a linear relationship between HB concentrations in saliva and plasma (r = 0.92). Mean salivary levels were 34 per cent of plasma levels. Salivary pH was constant throughout the experiment, 7.06 +/- 0.09. There was an inconsistent tendency of the saliva over plasma ratios to increase as a function of time. The percentage of protein binding calculated from saliva over plasma ratios was in reasonable agreement with in vitro data of equilibrium dialysis, 64.1 +/- 2.6 per cent and 65.9 +/- 0.8 per cent, respectively. The experiment was repeated in 4 subjects, and considerable intraindividual differences were shown to exist in saliva over plasma ratio, half-lives, and protein binding. It was concluded that HB elimination half-lives can relatively accurately be determined from salivary concentrations. Oral plasma clearance can only be estimated if the individual saliva over plasma ratios are known; this would require the taking of at least one blood sample during the experiment. When employing HB as a model substrate for drug metabolizing enzyme activity in vivo, the determination of its pharmacokinetic parameters, particularly oral plasma clearance as a reflection of cytochrome P-450 activity, cannot be achieved by taking saliva samples only.