EXPLORING THE USE OF COMPOSING STRATEGIES IN HIGH SCHOOL STUDENTS’ WRITING ACTIVITY PERFORMANCE INTRODUCTION This study explores the use of composing strategies within the performance of writing activities by a group of 10th graders at a public school in Monteria, Colombia. It specifically analyzes how the use of these strategies varies in regard to aspects such as the focus, target audience, topic and length of the writing activity. METHODOLOGY To do this, questionnaires, observations, diaries and videotaping were applied to the participants, enquiring on their perceptions of writing as a language learning skill, as well as their prior experience in writing texts in English. RESULTS These learners did not manifest special interest in learning how to write, but they considered writing a useful challenging skill and reported having previously performed in different writing activities. Among these activities, there was a diary which they had been writing since the beginning of the school year, of which some samples were collected. An analysis of the linguistic, discourse and pragmatic components of these sample journal writings was subsequently conducted with the purpose of establishing a baseline level of writing competence for the learners involved. It was found that those students whose writing was longer and more linguistically varied displayed higher incidence of grammatical inaccuracy. On the other hand, those who showed better accuracy tended to write shorter and less varied text. This observation suggested that, at the time of writing, learners may focus either on the grammatical accuracy of the text, or on the meaning conveyed by it. Besides, from the analysis of the sample journals, it was found that, among the most common strategies reflected on learners’ writing, there was textual reduction, topic avoidance and coinage of foreign terms. CONCLUSIONS Finally, participants’ real-time performance in writing activities was analyzed with the help of videotaping. In the first activity, a participant was told to write an anecdote. However, the activity was intentionally set up for him to experience some pressure (time was limited, no previous sharing was allowed and there was no access to reference resources, such as dictionaries). From his performance, it was observed that, when such limitations exist, learners tend to prioritize time as the key performance factor, which is reflected in shorter textual length, occurrence of local planning and mother language-based memory retrieving. In the second activity, a group of three participants were made to narrate a piece of news following a set of pictures in chronological order. In this case, they were granted enough planning time before writing, as well as access to dictionaries and direct help from the researcher, and untimed performance and revision time. It was concluded that, despite the less restrictive conditions, the participants had to cope with their lack of declarative knowledge on the writing topic by using local lexical planning and description-based generation of ideas. It was also found that planning before writing reduces actual writing performance and revision. The third and fourth activities were about describing an ideal partner and writing an autobiography. Here, the research focus was on the differences in strategy use and textual quality between a first draft and a final submitted draft. It was found that, at the time of revising their first draft, learners focused on specific linguistic and pragmatic devices, rather than on general discourse choices. The personal significance of the activity topic was associated to the length and eventual linguistic and pragmatic richness of the resulting text. Finally, it was observed that, when a model of a desirable final text is provided within the activity, learners’ choice of lexis and grammar can be influenced by it.