Oscar Wilde’s novel The Picture of Dorian Gray (1891c) offers an intriguingly sensuous read-ing experience. In the study reported here, two approaches within corpus stylistics were com-bined with cognitive stylistics and foregrounding theory, pursuing three aims. Firstly, the goal was to investigate stylistic characteristics of the novel via a bottom-up corpus-driven method-ological approach. Secondly, the colour motif was analysed in a top-down corpus-assisted approach. Thirdly, the two approaches were compared in terms of results yielded, and diffi-culties and advantages encountered. Throughout both analyses, the corpus methods yielded the quantitative data used as a stepping stone for the qualitative analysis, founded in cogni-tive stylistics and foregrounding theory. The corpus-driven analysis yielded innovative and rather dissimilar results, pointing towards multiple stylistic effects: the creation of dramatic effect; an effect of auditory stimulation; defamiliarization processes of beautiful entities; and a foregrounding of subjective intuitive feeling. The corpus-assisted analysis yielded focused results regarding the way in which the colour motif comes to be perceived as foregrounded. First of all, this analysis revealed that colour adjectives are used significantly more frequently in The Picture of Dorian Gray than in the three reference corpora consisting of Wilde’s works, Gothic literature, and literary works of Wilde’s contemporaries. It also revealed that the use of hyphenated compound adjectives is a unique characteristic of the novel and that a motif of possessions is attached to the colour motif, foregrounding dead objects in the text. Moreover, it was found that the novel’s style is poetic and rich in the sense that the textual level is embellished and heavily adorned, making for an aesthetically powerful reading expe-rience. The impression of a strong sensuous aspect was substantiated by the analysis, in which it was found that the text creates visual and tactile stimuli in the mind of the reader. In the comparison of the two approaches, it was found that the corpus-driven approach offers quite a few disadvantages, including a fragmented argumentation, disparate outcomes, and a de-tachment from the co-text. These issues have to be amended by the stylistician at some cost of time but resulting in an analysis with high validity and originality. In the comparison, it was also found that the corpus-assisted approach had a clear advantage in its intrinsic focus, making for a targeted and unified argumentation but at the cost of possible spontaneity and serendipity. Consequently, it can be argued that in stylistics a combination of methods can be beneficial, though no exact estimate of the optimal distribution of the approaches can be of-fered.