Conditions of low and high knowledge about the topic of a writing task were compared in terms of the time and cognitive effort allocated to writing processes. These processes were planning ideas, translating ideas into text, and reviewing ideas and text during document composition. Directed retrospection provided estimates of the time devoted to each process, and secondary task reaction times indexed the cognitive effort expended. Topic knowledge was manipulated by selecting subjects in Experiment 1 and by selecting topics in Experiment 2. The retrospection results indicated that both low-and high-knowledge writers intermixed planning, translating, and reviewing during all phases of composing. There was no evidence that low-and high-knowledge writers adopt different strategies for allocating processing time. About 50 % of writing time was devoted to translating throughout composition. From early to later phases of composing, the percentage of time devoted to planning decreased and that devoted to reviewing increased. The secondary task results showed that the degree of cognitive effort devoted to planning, translating, and reviewing depended on the task. Also, the high-knowledge writers expended less effort overall than did the low-knowledge writers; there was no difference in allocation strategy across planning, translating, and reviewing.Research on the writing process has increased markedly in recent years (Beach & Bridwell, 1984;de Beaugrande, 1980; Frederiksen & Dominic, 1981; Gregg & Steinberg, 1980;Nystrand, 1982). This interest is warranted given the importance of writing in work settings (Faigley & Miller, 1982;Odell & Goswami, 1984) and the concern raised by the ongoing National Assessment of Educational Progress that schools are not adequately preparing students in writing. Although research advances are being made, numerous basic questions remain, at best, only partially answered. One such question is how the knowledge of the writer affects the writing process. The importance of knowledge differences in other tasks, such as reading comprehension (Voss, Vesonder, & Spilich, 1980) and problem solving (Chi, Feltovich, & Glaser, 1981), suggests that this should be a key factor in writing.Writers differ in what they know about their language, audience, and topic (Applebee, 1982). The present studies examined how knowledge of the topic being written about affects the amount of processing time and cognitive effort allocated to various writing processes. Although linguistic and audience knowledge also deserve exploration, they fall beyond the scope of this initial investigation. Processing time and cognitive effort were selected because, as will be discussed below, they concern two prominent features of writing based on current theory and empirical findings.The author thanks the anonymous reviewers for the useful suggestions they provided on an earlier version of this manuscript. Requests for reprints should be addressed to Ronald T. Kellogg, Department of Psychology, University of Missouri-Rolla, Rolla, M...