Saliva is an attractive source for oral microbial detection and quantification since sampling is non-invasive and rapid. Objectives: To determine whether different saliva preparation methods or preservation time periods affect DNA stability. Methods: Saliva samples from 4 healthy adult volunteers were processed to obtain 3 different preparations: whole saliva, and after centrifugation pellet and supernatant. Purified DNA (MasterPure (TM)) from each sample was divided into 4 aliquots, one for immediate analysis and 3 (stored at -80 degrees C) for later analyses after 1 week and 2 and 6 months. DNA concentrations and qPCR based quantities of Porphyromonas gingivalis, Prevotella intermedia, Parvimonas micra, Fusobacterium nucleatum, Filifactor alocis and Streptococcus mutans were determined. Results: DNA concentration did not decrease (P > 0.05) during the 6-month period in any sample. Mean (SE) DNA concentrations (ng/mu l) in whole saliva were 152.2 (51.2) and 147.8 (50) at day 0 and 6 months, respectively. Similarly, the values for pellet were 134.9 (42.5) and 133.6 (42.9), and for supernatant, 11 (1.9) and 8.9 (2.3), the difference being significant (P < 0.001) between supernatant and whole saliva or pellet. The quantities of most bacterial species found at day 0 remained stable over the 6-month period in all saliva preparations. In supernatant, species quantities were lower (P < 0.05) than in whole saliva or pellet. Conclusions: DNA concentrations were comparable between whole saliva and pellet, suggesting that either of them can be used for DNA-based analyses. Our results also demonstrated that DNA extracted from saliva can be preserved at -80 degrees C for at least 6 months without decrease in DNA concentration.