Three different methods for detection and count of Streptococcus mutans in saliva have been compared with a microbiological method in order to estimate their efficacy and practicality. The study has been carried out on fifty children of six-eleven years old. The first method (stamp method) derives from that of Kohler and Bratthall, improved on the sampling, without saliva stimulation and, hence, more rapid. Furthermore, using a tongue depressor with a shovel-like extremity, we can either restrict the area where the grown colonies must be counted, avoiding any mistake, and stamp the spatula in the centre of the plate, not in the border, as the original method described, pressing it with the same force in every part, so that almost all the saliva collected can be plated on the Streptococcus mutans selective medium. As stated by this method, a tongue depressor is pressed first on the subject's tongue, then on a plate containing the medium. After incubation, the Streptococcus mutans colonies are counted on the plate's area limited by the stamp of the extremity of the spatula. According to Emilson we can also discriminate Streptococcus mutans and Streptococcus sobrinus colonies. The second method is a "strip mutans test" (Dentocult): saliva is stimulated making the subject chew a piece of paraffin and then it is collected passing a plastic spatula for ten times on the child's tongue. The spatula is introduced into a tube containing a liquid medium selective for Streptococcus mutans and then incubated.(ABSTRACT TRUNCATED AT 250 WORDS)