Building on Robinson’s (2001, 2003) Cognition Hypothesis, Skehan’s (1998) Limited Attentional Capacity Model, and Kellog’s (1986) model of writing, this study examined the effect of task planning on the fluency, accuracy, and complexity of 60 Iranian EFL learners’ argumentative and narrative writings under different planning conditions. A quasi-experimental design with three levels of planning conditions (pre-task planning, within-task planning and no-planning) and different time constraints was used. Measures of fluency, accuracy, and complexity were used to evaluate the quality of the participants’ written productions. The results of a series of one way Multivariate Analyses of Variance (MANOVAs) indicated that increasing task complexity, through task planning conditions and also different tasks produced significant differences among the groups in fluency and accuracy; however, with regard to syntactic complexity, both tasks provided similar results among the groups. The findings add support to the view that selecting an appropriate task with appropriate task-based implementation conditions can induce language learners to increase, accuracy and fluency, but not syntactic complexity of their output. Pedagogical implications are discussed and suggestions for further studies are made.
