Coherence plays a central role when readers construct meaning from a text. Previous research has shown how coherence marking affects text processing and representation. However, this effect seems to depend on reader's prior knowledge of the text content: Low knowledge readers benefit from coherence marking, whereas high knowledge readers benefit from a more implicit text (McNamara & Kintsch, 1996). Because this interaction was not consistently found in previous research, this article takes a closer look at the operationalization of the experimental variables: coherence marking, prior knowledge, and text comprehension. Also, this article compares the effect on both informative and persuasive texts. Results indicate that linguistic marking of coherence indeed interacts with prior knowledge in the informative genre, but not in the persuasive genre. (Contains 2 tables, 2 figures and 3 footnotes.)