The present paper investigates linguistic norm-adherence in Belgian Dutch written and audiovisual translation. More particularly, it is measured to what extent language use in subtitles, in comparison to regular written translations and non-translations, conforms to explicit linguistic norms. Additionally, we analyze which effect different contextual parameters have on the extent of norm-adherence in Belgian Dutch subtitles. We use the Dutch Parallel Corpus and the SoNaR Corpus, and we analyze the data by means of profile-based correspondence analysis, yielding a visualization of norm-adherence distances between the different translation modes and non-translations. The results reveal that the parameters speaker type and source language significantly affect the degree of linguistic normadherence, whereas program genre has no influence. It is also shown that norm-adherence in subtitles holds a middle position between written translations and non-translations, which is explained in terms of target audience and communicative risk.