Examined the relationship between reading behavior and verbal ability to test the hypothesis that verbal ability influences reading time components and thus has an effect on test performance. 119 students from German secondary schools aged between 13 and 17 years read texts and answered multiple-choice questions with or without access to the text. Verbal and reading ability measures were completed and resource allocation at the word, sentence, and text level was measured. The relationships between reading time components, resource allocation, verbal ability, and comprehension were examined. Results confirmed a systematic relationship between resource allocation and verbal ability and test performance. Test performance was better when the text was available. Students scoring higher in verbal ability and comprehension tended to perform better in both test situations. In the without-text condition, reading time components related to high-level integration processing were more important. In conclusion, in test situations where the text is not available, the level of mental representation formed by the reader has an important effect on comprehension test performance. (ZPID).