In the year 806 CE the Japanese monk Kūkai returned from a journey to China and brought a large amount of visual artefacts with him. Commentators have wondered since what role these visual media play in Kūkai’s Buddhist thought. It has been speculated that the art works show that Kūkai values visual media higher when it comes to transmitting the teaching of the Buddha. Proponents of this view usually refer to a single passage from Kūkai’s writings to warrant their interpretation. By analysing the respective passage in detail and showing how it connects to Kūkai’s other writings, this article argues that Kūkai did not prefer the visual to the verbal in transmitting the dharma. Mandalas certainly play an important role in Kūkai’s thought, but their role differs from what these modern interpreters suppose: first, when Kūkai speaks about ‘mandalas’ he often does not refer to paintings, but to the structure of reality or to ritual procedures. Second, mandala paintings have an ambiguous role in esoteric ritual, because they were added rather late in the development of esoteric ritual. For Kūkai they serve primarily as storyboards for ritual performance. Third, the first glance at a mandala is an important moment during esoteric initiations, but it is only the beginning of a rigorous training. Moreover, the crucial moment in esoteric ritual is the union of the practitioner with the deity; glancing at the mandala has no role to play in this mystic union. Fourth, mandala paintings can be used, according to Kūkai, to reveal the deeper structure of texts, but in this role they are not superior to the written medium but rather play a helping role. Fifth, Kūkai believes that texts as well as paintings can be misleading whenever they are taken as representations of a rigid structure of reality. In Kūkai’s eyes, the visual cannot, therefore, solve the problem how the Buddha can transmit his dharma.