Grounded cognition assumes that language is understood using simulations in different modalities. Evidence for this assumption mainly stems from studies using concrete concepts (e.g., action related language makes use of the motor system, concepts related to power are represented as vertical differences in visual space). However, less evidence for grounding exists for abstract concepts, which are assumed to be grounded via metaphors associated with them. In the present study, we investigated visual grounding of abstract concepts loosely related to power or the exercise of power in different ways. As stimulus material, we chose pairs of concepts, e.g., democracy and dictatorship. Participants were presented each word separately and asked to create a visual image in their mind. Then they were asked to rate images on several aspects, e.g., how spontaneous the image is, how colorful it is. Afterwards they were asked to draw a sketch of the image. Results showed systematic differences between more and less powerfully concepts in ratings and drawings, though results were not consistent for all pairs of concepts. Those inconsistencies were however related to the specific content of the concepts. For instance, images of more powerful concepts (e.g., wealth vs. poverty, wisdom vs. stupidity) were rated as more colorful, with the exception that democracy was rated as more colorful than dictatorship. This may be explained by the conception that democracies allow for more diversity and are thus, metaphorically, more colorful. In conclusion, abstract concepts related to power are grounded in the visual modality.