This study offers a detailed and systematic account of Dutch orthography and its relation to the linguistic system. It reveals aspects of the spelling system that were ignored or left implicit until now, and integrates them with what was already known from prescriptive and descriptive accounts. The most important insight this study yields is that the Dutch spelling system consists of two distinct components: phoneme-to-grapheme conversion rules and autonomous spelling rules. Thus, the computation of spelling from sound representations is a two-step process. The investigation also shows that the phoneme-to-grapheme conversion rules for non-native words should be distinguished from the rules for native words. The autonomous spelling rules apply to both sets of lexical items alike, however. Dutch spelling rules were modelled by means of a computer programme and applied to the sound representations of Dutch words.