Objective: The objective of this descriptive and cross-sectional study was to classify and to determine the sort of the students spelling errors, and to study the Persian forms orthographic processing on the basis of linguistic and non-linguistic information in children. Method: The sample was 85 students. The Garman’s (1996) classification was used. This classification suggests that while spelling, children use their memory as well as their linguistic knowledge that might be orthographic, morphological, syntactic or semantic. An orthographic test comprising 246 “goal lexemes and phrases” and 20 dictations from the beginning, the middle and the end of the academic year was the instrument o f the research. Results: the findings manifested that the most spelling error was for the letter level, and the least related to symbol level (e.g., gemination). Regarding the kind of errors, substitutions had the greatest percent. Besides, displacement, repetition and mixed forms had the least percent and were mostly nonlinguistic, and one can assume them as slips of the pen, like speech slips. At the language levels, the most percent of errors was for “lexeme” and the least was for the “syllable”. Conclusion: Most dictation errors were linguistic and originated from the Persian writing system. Furthermore, the findings stress that children enjoy both linguistic information and the visual memory. Processing the orthographic forms which are produced by using linguistic information is easier for them than processing the forms produced by using nonlinguistic (visual) information.