We studied whether honeybees can distinguish face-like configurations by using standardized stimuli commonly employed in primate and human visual research. Furthermore, we studied whether, irrespective of their capacity to distinguish between face-like stimuli, bees learn to classify visual stimuli built up of the same elements in face-like versus non-face-like categories. We showed that bees succeeded in discriminating both face-like and non-face-like stimuli and categorized appropriately novel stimuli in these two classes. To this end, they used configural information and not just isolated features or low-level cues. Bees looked for a specific configuration in which each feature had to be located in an appropriate spatial relationship with respect to the others, thus showing sensitivity for first-order relationships between features. Although faces are biologically irrelevant stimuli for bees, the fact that they were able to integrate visual features into complex representations suggests that face-like stimulus categorization can occur even in the absence of brain regions specialized in face processing.