This study investigates the impact of word processors on students' writing. It compares two Year 8 classes, one using pens for composing, the other, computers, within the context of an all-girls' school. A major emphasis of the research is the computer's impact on the quality of the texts the students produced. As part of the pretest-posttest design, the students responded to a questionnaire, to a writing tool preference inventory, and were required to produce texts which represented three genres: narrative; argument; and report. The texts were evaluated for productivity, syntactic complexity, precision, global quality and quality as determined by analytic measures. In order to examine the teaching-learning context, the researcher adopted the role of participant/observer, attending all lessons and recording observations, conversations, interactions and critical events. The findings suggested that the computer classroom was more student centred, less teacher dominated and more work focused. Overall the atmosphere was more cooperative, collaborative and interactive. The examination of the writing processes associated with particular tools revealed that composing behaviours varied according to the genre task rather than according to the influence of the writing tool. The preference data indicated that the computer students had a strong preference for composing narrative with pen and argument with word processors. The investigation of the quality of the texts produced with different tools offered strong evidence of the efficacy of word processing in the promotion of quality in argument and report and weaker evidence for narrative. These findings suggested that word processors are indeed effective as writing tools. Further, the findings suggested that word processors are useful for the genres with which junior- secondary students are less familiar, less confident and less competent, but which are recognised as instrumental in determining success at school as well as in the wider sociocultural context beyond school.