The aim of this study is to compare reading comprehension levels of two groups of students--an experimental and a control group--who read the same texts in course books using either augmented reality technique or traditional reading technique. Barrett's taxonomy, which proposes a classification of reading comprehension, consisting of literal comprehension, reorganization, inferential comprehension, evaluation and appreciation, was used to determine the reading comprehension levels of the students. The texts based on the augmented reality were produced with an augmented reality application developed by the authors using the Kotlin programming language. The participants of the study were 120 students aged 11-12 who attend a secondary school in Antalya (Turkey). In the pre-test phase, both the experimental and control groups read the same texts from the course book and answered the comprehension questions. In the post-test, the experimental group read the augmented reality texts using tablets while the control group read the texts in the traditional way. Then both groups answered comprehension questions. The results of the pre-test revealed that there was no significant difference between the two groups in terms of their total comprehension scores and their reading comprehension levels. The post-test results indicated that the experimental group's overall comprehension scores were much higher than those of the control group. In terms of reading comprehension levels, no difference was found between the two groups in terms of literal comprehension levels. However, the experimental group is found to be much more successful in terms of the reorganization, inferential comprehension, evaluation and appreciation. These findings suggest that the augmented reality content may contribute to the students' reading comprehension performances.