The aim of this study was to evaluate the influence of caries-detecting dyes on the microleakage of adhesive materials. Sixty cubic class V cavities were prepared on buccal and lingual surfaces of 30 human third molars. Coronal margins were located in enamel and gingival margins in cementum. The teeth were randomly divided into six groups of ten restorations each. Cavities were restored with an adhesive system (Single Bond, 3M ESPE, St. Paul, Minn., USA), a compomer (F2000, 3M ESPE), or a composite resin (Z100, 3M ESPE) according to the manufacturer's directions. Acid red dye (Seek, Ultradent, South Jordan, Ut., USA) and basic fuchsin dye (Vide Cárie, Inodon, Porto Alegre, Brazil) were tested. Control groups were prepared without the use of dyes. After 7 days of storage in distilled water, the restorations were polished and the teeth were subjected to thermal cycling followed by immersion in 2% methylene blue. The teeth were sectioned, and microleakage scores were evaluated under magnification (40x). Data were submitted to statistical analysis using the nonparametric Kruskal-Wallis test. A statistically significant difference ( P<0.05) in microleakage was found between the materials in cementum (Z100>F2000) but not in enamel. Control and experimental groups using dyes showed similar results. It was concluded that dyes for caries detection did not increase microleakage of the adhesive materials tested.