A study examined the composing processes of eight college freshman writers and the contexts for their writing when they composed in natural settings without narrow time constraints. Subjects, both traditional freshmen and adult learners with varying degrees of writing skill, were assigned two research tasks during which they kept tape-recordings and logs of their composing sessions and processes, as well as all their written products. The first task was a carefully subdivided persuasive writing assignment, while the second task was a "free choice" assignment for which students had to invent the writing problem as well as its solution. The results indicated that students varied considerably in the frequency of their composing sessions, and that time was the single most important external factor affecting students' composing processes. The recursive nature of the composing process was supported by the results. Most subjects were capable of successfully drafting short papers without having any written plan for their drafts, suggesting that traditional "outlining " methods are simply not necessary. The persuasive task elicited more varied process strategies for generating ideas, engaging students in use of analogy, conditional argumentation, hypothetical scripts, and other imaginative strategies. Students generated more ideas at all levels for the problem paper, and also more written plans; despite this, however, final drafts of these papers were an average of six sentences shorter, suggesting that translating ideas into words was harder for the persuasive task. The results have implications for teachers that involve the need to set up supportive writing environments and to establish instructional objectives and make assignments that draw on students' own experience. Implications for the direction of writing research in the areas of composing contexts, composing processes, and task variables are also discussed. (HTH)