The goal of this in vitro study was to identify the most suitable dye for endodontic dye leakage studies, which could be a further step towards standardisation. The root canals of 70 extracted, single-rooted human adult teeth were enlarged to apical size 50 using hand instruments. The teeth were divided into seven groups (n = 10 each), and all root canals were completely filled by injection with one of the following dyes: methylene blue 0.5% and 5%, blue ink, black ink, eosin 5%, basic fuchsin 0.5% and drawing ink. Transverse root sections from the coronal, middle and apical part of the roots were examined, and the percentage of the dentine penetrated by dye was evaluated by software-supported light microscopy. In addition, the range of particle size of drawing ink particles was evaluated. There were conspicuous differences in the relative dye penetration into the root dentine and the penetration behaviour in the different root sections (two-way ANOVA, both p < 0.0001). One dye (drawing ink) penetrated less into the root dentine compared with all the others (p <0.0001). The particle size of this agent (0.1-2 microm) corresponds best with the size range of a representative selection of 21 species of pathogenic endodontic bacteria. Compared to the other dyes tested, drawing ink appears to be superior for use in endodontic dye leakage studies. The penetration behaviour into the root dentine of all the other dyes tested might be one factor that limits the applicability of these dyes in dye leakage studies.