Background. Mouth characteristics are considered important yet restraining factors for food acquisition in fishes. Nevertheless, their relations with total length (TL) are not thoroughly examined. In the present study mouth characteristics were estimated for 61 fish species from the northern Aegean Sea and their relations with TL were established, and the relation of mouth area with feeding habits was tested. Materials and methods. TL, horizontal mouth opening (HMO), and vertical mouth opening (VMO) were measured in 61 species, and mouth area MA was estimated. Relations between mouth characteristics and TL were established using power regression, and the validity of the power model was tested, using Student’s t-test. In order to explore whether there is a grouping of species based on their feeding habits, the MA–TL regressions of all species were plotted together and the general regression lines per functional trophic group were compared using analysis of variance. Finally, the relative MA was estimated and related to trophic level. Results. All relations examined were statistically significant (P < 0.05), with only one exception (VMO–TL relation in Dentex dentex). The allometric model (for HMO–TL, MA–TL, and VMO–TL) was valid for the majority of species (52, 47, and 49 species, respectively), with positive allometry prevailing in the HMO–TL and MA–TL relations, and negative allometry in the VMO–TL relations. The analyses employed revealed that MA, for the same TL, increases faster in carnivores than in omnivores, and that in general carnivorous species tend to have larger mouths than omnivorous ones. Finally, there is a strong positive relation between relative MA and trophic level (P < 0.01). Conclusion. There is a strong relation of mouth characteristics to body length and feeding habits. These relations and the way that MA increases in relation to feeding habits could be attributed to structural changes in order for growing fish to meet their increased energetic demands. Establishing such relations is essential for estimating trophic levels of species, which are key parameters for ecosystem-based management models, when data on species’ feeding habits are not available.