Purpose: First, to elucidate the cause of the appearance of "white lines" in Class V restorations as seen under the light microscope and their enhancement when using a fluorescent microscope (FM); second, to compare the results of the FM quantitative marginal analysis to those of the SEM quantitative marginal analysis. Materials and Methods: Thirty-two standardized Class V fillings, with half of the preparation in dentin and half in enamel, were placed in 16 human caries-free premolars on the buccal and lingual surface using 4 adhesive systems (Clearfil SE, Syntac, 2 experimental self-etching adhesives) and the composite Tetric Ceram. All teeth were connected to a device containing horse serum to simulate dentinal fluid and subjected to thermomechanical loading (1,200,000 cycles at 49 N, 3000 cycles at 5°C/55°C). The restoration margins of 24 fillings were directly evaluated with FM - after the fabrication of replicas - and with SEM. Eight filllings were used to elucidate the cause of the fluorescent signal, by examining them with FM and different filters under dry and moist conditions, and also with dark field microscopy and CLSM. Further, two Class V fillings with the same composite but without the fluorescent substances were made and evaluated. Selected specimens were cut in the bucco-oral direction to analyze the restorative interface on section replicas with SEM. To detect differences between FM and SEM, the Wilcoxon test was performed (p 0.05). To evaluate the degree of agreement between the two evaluation methods, a linear regression analysis was performed and the Spearman correlation coeffcient was calculated. Results: The evaluation of Class V restorative margins by SEM, dark field microscopy, and CLSM revealed that the phenomenon seems to be an optical effect caused by the debonding of the composite from the underlying substrate creating another refractive environment when air penetrates into the gap. The effect depends on the geometry of the cavity design, the absence of moisture, and the use of a fluorescent composite material. The light source that produced the best visible results covered both ultraviolet and the blue part of white light (350 to 460 nm), producing light green lines and green areas. Those green or white areas as seen with the light microscope were related to gaps in the interface between tooth substance and composite. With the exception of one of the experimental self-etching adhesives, the mean percentage of continuous margin of the 4 groups showed no statistically signifcant dfference for the two test methods. Linear regression analysis revealed a good correlation between the two test methods for dentin margins and a moderate correlation for enamel margins. Conclusion: Fluorescence microscopy seems to be a useful tool for evaluating dentinal and enamel margins of Class V restorations in vitro. Palabras clave: restoration margin, fluorescence microscopy, SEM, autofluorescence, white lines