Many studies have demonstrated the facilitating role of rhetorical devices in text comprehension, but there are also studies where rhetorical devices have not shown such effect. The present study sets out to explore whether readers' knowledge of rhetorical devices (that is, rhetorical competence) moderates their effectiveness beyond general comprehension skills and, consequently, whether rhetorical competence may be considered a component skill of reading comprehension. 192 sixth- to seventh-grade students were assessed on rhetorical competence and were required to read a difficult marked text with specific rhetorical devices (a refutation, an objective, and four organizational signals) or the same text without them. After reading, students produced a summary in order to obtain three dependent variables: main ideas (as a measure of participants' ability to select relevant information from the text), causal links between them (as an indicator of participants' ability to grasp the logical structure of the text and to organize its ideas), and the combination of main ideas plus causal links (as an indicator of participants' global comprehension of the text). Analyses controlling for general comprehension skills and other important variables (working memory, prior knowledge, decoding) demonstrated that: (a) readers of the marked text scored higher in terms of all dependent variables, and (b) rhetorical competence level moderated the effect of rhetorical devices on the composite measure (main ideas plus causal links) and on the organization of the summary by means of causal links.