The present study was designed to investigate the composing processes or strategies used by EFL writers at Sultan Qaboos University (SQU). More specifically, it addressed three major research questions which are: 1. What are the similarities and differences between the composing processes of skilled, semi-skilled and unskilled EFL Omani students? 2.To what extent are these processes related to the learners' overall English proficiency as determined by their English GPA's? How do these processes vary according to instructional level? The sample consisted of 165 students majoring in English as a Foreign Language in the College of Education and Islamic Sciences at SQU. These subjects were drawn from three instructional levels (Semesters 7, 5 and 3). The instruments used were: (1) written protocols which consisted of stu dents' outlines, drafts, and final compositions; (2) written self-reports ob tained from all of the subjects in the sample. The students' final products were then evaluated by two experienced professors on the basis of holistic scoring. A mean grade of the two evaluations was calculated in order to see in which group each writer fell: the skilled group (a grade of B and above) or the semi-skilled (a grade of B- to C) or the unskilled groups (a grade of C and below). The written protocols as well as the self-reports were analyzed togeth er with the purpose of identifying the presence or absence of some pre determined composing strategies. The frequency of occurrence of some components of the writing process (e.g. number of changes and number of pages written) was important, so the times of occurrence were counted to find out whether the skilled, the semi-skilled and the unskilled writers dif fered with respect to that. The reliability of the coding was established through cross-validation in which a random sample of 25 essays with their self-reports were analyzed by another judge (in addition to the researcher). The two analyses were then subjected to the correlation technique. The value of the correlation was .78. The main results of our study were as follows: 1) Some composing strategies seem to be more important than others in differentiating between the groups of skilled and unskilled writers. Some of these are: audience consideration, purpose consideration, concern with content in the revision stage, coherence, number of strategies reported (awareness of the strategies used) organization and development of ideas. The composing process factors were not found to be greatly related to the subjects' overall English proficiency as determined by their GPAs. Only three of these factors correlated significantly with this measure of academic achievement in English. These are: use of complex and varied vocabulary, proof reading while writing and the number of changes made while writing. The instructional level was found to have main effect in the use of certain writing strategies. The three instructional levels differed significantly on a number of writing strategies. Seventh Semester students, for example, had a better conception of audience, better organization and development of ideas than both Semester 5 and Semester 3 students. The differences suggest that, overall, qualitative changes occur in the students'writing process as they move from one instructional level to the other. Among the research implications suggested by the results of this study are the following: (a) the need to establish whether the strategies that differentiate the skilled and the less skilled writers are consistent among different writing tasks, (b) the need to investigate the effect of some back ground factors (e.g. writing and reading ability in the Li) on the composing processes of different writing proficiency groups. With regard to the peda gogical implications, it was recommend that more attention could be given to the teaching of the composing processes that were found to differentiate the skilled from the less skilled writers.