Background. How indispensable are examples and main ideas in study texts? In research into comprehension of expository texts examples are sometimes considered as cognitive support, sometimes as seductive details. According to the cognitivist view, text comprehension is based on main ideas, whereas the constructivist view holds that examples are the basis of understanding. Aims. This study explored how text comprehension is influenced by main ideas and examples in study texts, in relation to Vermunt's (1992, 1998) 'concrete elaboration' learning style. In Experiment 1, concrete texts with many examples were compared with abstract texts with many main ideas. In Experiment 2, idea-oriented texts, in which main ideas preceded examples, were compared with example-oriented texts, in which examples preceded main ideas. Samples. In both experiments, undergraduate social sciences students studied various versions of an introductory text on educational psychology. Methods. The text contained sections with a varying number of relevant and irrelevant examples and with or without a main idea (Experiment I), and sections with a main idea followed by examples, sections with a main idea without examples, or sections with examples followed by a main idea (Experiment 2). After studying the text, students completed a verbatim recognition test and an explanation test. Results. Best results were obtained after studying sections with a main idea and two or five examples. Irrelevant examples were detrimental to understanding. Students used examples to construct knowledge or to activate prior knowledge. Students with a strong habit of concrete elaboration used main ideas to recollect episodes of personal experience, This may interfere with understanding underlying concepts and principles by relating main ideas to examples in the text, Students with a low score on the concrete elaboration scale were sensitive to the presence of examples in the study text. Conclusion. In expository texts, examples are indispensable. The findings suggest that main ideas are useful, and, in order to prevent interference effects, the more so when they are put at the end of sections.