The potential of hydrogen peroxide to reduce the levels of salivary thiol precursors of oral malodor was investigated in-vitro and in-vivo. In both cases the concentration of thiol groups was determined colorimetrically by quantitative reaction with 4,4'-bis (dimethylamino) diphenyl carbinol. Addition of volumes of hydrogen peroxide solution (containing between 0.18 and 0.90 mmol) to premeasured aliquots of saliva in-vitro, resulted in reductions in salivary thiol levels of between 53% and 75% compared to controls. This positive indication prompted an in-vivo investigation. The efficacy of a fluoride-containing test toothpaste also containing 0.67% hydrogen peroxide and 5.48% sodium bicarbonate was evaluated in a crossover study using ten male and female subjects (non-smokers). All subjects used the test product and a control fluoride dentifrice, in a random order. For the duration of the study subjects used a standard silica based toothpaste containing 1500 ppm F (as sodium monofluorophosphate) exclusively for their normal oral hygiene. On each sampling morning they refrained from oral hygiene and eating and drinking on rising. At the test facility they generated a background saliva sample stimulated by chewing unflavored, unsweetened gum. Subjects brushed for 1 minute with 1.50 (+/- 0.05) g test or control paste and generated another saliva sample as before, 30 minutes after product application. Using the same analytical procedures the mean (+/- SEM) percent reduction in salivary thiol levels post treatment compared to baseline was found to be 59.0 (+/- 7.0)% for the test product compared with 12.5 (+/- 5.2)% for the fluoride control paste.(ABSTRACT TRUNCATED AT 250 WORDS)