AbstractBacteriologic and clinical studies suggest that sealants can be used therapeutically on incipient caries of the enamel. The present study was designed explore this approach further by determining whether there is a difference in retention rates on sound or carious tooth surfaces sealed for the first time as well as resealed surfaces. A total of 1766 teeth were sealed for subjects 12–14 yr of age of which 120 were judged to be carious. Mouths were divided sagittally on a random basis and an ultraviolet light activated resin with a filler or an autopolymerizing resin without a filler was placed on contralateral surfaces of all molar and premolar teeth. The effects of sealant type, caries status, arch and tooth type on retention over a 2‐yr period were tested using Mantel‐Haenszel statistics. At the end of 1 yr, 88% of the teeth sealed with the autopolymerizing resin were judged to be completely intact compared to 82% of the teeth filled with the ultraviolet light resin. At the end of the second year the retention rates were 84% and 75% respectively. The retention rates for carious and sound teeth at the end of both years were almost identical. Overall, premolars had a 15% more favorable retention rate than molars. There were minimal differences between arches although maxillary premolars had approximately a 5% higher retention rate than mandibular premolars. The reverse was true for molar teeth. Retention of sealants at the end of the second year (resealed after 1 yr) was 4% less than comparable teeth sealed init