Abstarct: The present study seeks to concentrate on their disability to choose appropriate words to present their intention and meaning. To approach this problem, the present study sought to analyze the Iranian writing errors baxsed on frxame semantic and also recognize whether a teaching method drawing on frxame semantic can have any significant effects on the Iranian writing improvement. One hundred English Language Learners, from 16 to 36 years old, were asked to write one composition. The instrument was a composition for the purpose of gathering data. After introducing the task, students were asked to write about the topic, with the presupposition that this is part of their final exam so that they could do their best and consider it seriously. After collecting the data, the participants’ compositions were analyzed baxsed on frxame semantics principles to find more about the lexical selection errors made by students. The data were analyzed following some steps of error analysis model drawing on the principles of frxame semantics. The findings identified 12 types of errors the most frequent of which was word choice, while the least frequent type of error was Conditional Sentences. The most frequent category of error was selection including 31% of all errors. As the second top frequent category, addition was 26% of all error closely followed by ordering 24% of all errors. As pedagogical implications, EFL teachers are advised to consider learners’ errors as good indicators of their learning problems so that they can help them how to deal with their errors.