A total of 106 molars with occlusal caries were sealed with an autopolymerizing sealant. Samples of the carious dentin were obtained 1, 3, 7, 14 and 28 d, and 2, 4, 6, and 12 months after sealant placement. The samples were cultured on enriched nonselective medium, a streptococcal medium, a Streptococcus mutans medium, and a medium for acidogenic microorganisms. The bacterial counts were compared with control samples obtained from nonsealed occlusal carious lesions, half of which had acid conditioner applied to the occlusal surface for 1 min prior to sampling. The mean total viable counts decreased by approximately 99.9% during the 1-year study period, from 925.1 x 10(4) CFU/mg of sample to 0.9 x 10(4) CFU/mg of sample. Statistically, there was a very strong linear relationship (P less than 0.001) between the reduction in bacterial counts and time. A similar reduction was observed for the total streptococcal count and the S. mutans count. The etching procedure itself reduced the number of cultivable organisms by about 75%. Sealant material that was polymerized directly in culture media did not result in any inhibition of bacterial growth. Although slight fluctuations in the relative proportion of the microflora were observed at the different sampling intervals, there were no significant changes in the relative distribution of bacterial types with time.