Labile reservoirs are important in maintaining ion concentrations in oral fluids, especially after a fluoride dentifrice application, where a persistent increase in fluid fluoride can mitigate or reverse caries progression. In this study, the effect of experimental and conventional fluoride rinses on the in vitro and in vivo water-induced release of fluoride, calcium, phosphate, acetate and hydrogen ions from oral reservoirs was examined. At the start of each experiment, 13 subjects rinsed either with a conventional 228-ppm fluoride NaF rinse, a 228-ppm fluoride controlled-release rinse (CR rinse) or received no rinse. Sixty minutes later upper and lower molar plaque samples and 1-min saliva samples were collected. The subjects then rinsed with deionized water for 1 min, and 7 min later, a second set of samples was collected (in vivo study). Plaque fluid and clarified saliva were then recovered from samples by centrifugation, and the remaining plaque mass was sequentially extracted with water and acid to measure the water-extracted and total whole-plaque fluoride (in vitro study). All the samples were analyzed using microtechniques for pH, free calcium, phosphate, organic acids (plaque fluid) and fluoride (plaque fluid, centrifuged saliva and plaque extracts). Results showed that in vivo water rinsing decreased acetate and phosphate in plaque fluid, and fluoride in plaque fluid and saliva, but had no effect on plaque fluid pH. In vivo water rinsing, however, increased plaque fluid free calcium, apparently due to water-induced loss of calcium-binding ions. Water- or fluoride-rinse-induced changes in plaque fluid concentration were greater at the lower molar site, suggesting that rinse pooling may influence ion distribution. Before the water rinse, plaque fluid, saliva and whole-plaque total fluoride values were 1.7, 2 and 4 times higher after the CR rinse compared to the NaF rinse. Furthermore, the CR rinse deposited approximately 11 times more water-extracted fluoride compared to the NaF rinse, suggesting a 'more efficient' precipitation of 'labile' or 'loosely bound fluoride'. The results presented here, and in previous studies, suggest the possibility of formulating effective fluoride dentifrices with a lower fluoride content than is currently in use. [ABSTRACT FROM AUTHOR]