AIM: To correlate the tray technique for isolation and counting of Streptococcus mutans over dental restoration with the toothpick technique. METHODS: Forty 18-50-year-old patients of both genders with high cariogenic risk were selected. Inclusion criteria were: occlusal restoration evaluated as alpha, according to the Ryge's criteria (examiner' Cohen-Kappa 0.8) and 2 years old maximum age. Each patient provided a saliva sample (control group S) and two samples of dental biofilm from amalgam or resin-based composite restorations in the same patient, one obtained with the toothpick technique (group TT), and the other from homologous tooth restoration with the tray technique (group TR). This method involves only a direct impression on the occlusal surfaces by a section of a fluoride tray with trypticase yeast extract cysteine sucrose with bacitracin (TYCSB) agar previously prepared and solidified. The samples were incubated at 37 ºC for 48 h and the colony-forming units (CFU) were counted. The results were statistically analyzed with SPSS software using the linear regression method at 95% confidence level. RESULTS: TR was able to isolate S. mutans, and there was a significant positive correlation (r=0.95) with TT. There was no significant correlation between TR and control group S (r=0.47). CONCLUSIONS: TR was able to isolate and count S. mutans from the dental biofilm of dental restorations with a significant correlation with the TT group.