Objective: This study aimed at analyzing and comparing the written narrative structures of fourth- and fifth-grade students with and without learning disabilities in Tehran. Method: The present study was descriptive-analytic. The statistical population comprised 10-12 year-old male students with and without learning disabilities in Tehran. The sample consisted of 20 male students with learning disabilities and 20 normal students (20 in the 4^th and 20 in the 5^th grade) selected using convenience sampling. A researcher-made reading test (a picture story) was used to elicit written narratives. The students were asked to rewrite the story after reading it. Their writings were assessed in terms of the number of words, mean length of utterance (MLU), and the complexity of the structure on macro- (character, setting, internal response, plan, action, and consequence) and micro- (coordinating and subordinating conjunctions, mental and verbal processes, and extended noun phrases) structure levels. Data were analyzed using independent samples t-test. Results: The normal students received 195 and 143 points, while the students with learning disabilities received 170 and 116 points on macro- and micro-structure levels. Both groups had better results on the macro-structure level. They both gained the highest score in character and the lowest score in plan on the macro-structure, and the highest scores in mental and verbal processes and the lowest score in extended noun phrases on the micro-structure level. The difference between the students’ scores in subordinating conjunctions and action was significant (p≤0.05). Also, the difference between the two groups was significant in terms of the number of words and MLU (p ≤0.05), and normal students produced longer narratives. Conclusion: The group with learning disabilities had a weaker performance on all linguistic categories, except for character. This difference was especially evident in terms of coordinating and subordinating conjunctions, action, plan, and extended noun phrases.