Learning to write correctly, besides the learning of fluent speaking a foreign language, as two important productive skills had been under a great attention from the beginning of the research over the English language teaching and learning. Errors and mistakes, as inevitable parts of learning a 2nd/FL have also paid much attention, though sometimes considered as a sign of progress, and sometimes as sins to be stopped. However, it requires the ESL/EFL teachers to provide the learners with appropriate written Corrective Feedback (CF) to prepare learners with desirable written Grammatical Accuracy (GA). Different written CF forms are available out of which this study investigates the effect of direct and indirect forms of it comparatively, on Iranian EFL learners' GA performance in sentence completion exercises (SCE). Fulfilling the requirements of the study, 83 homogenous participants in three intact classes were selected from a language school in Tabriz, Iran. Two direct and indirect written CF experimental groups, as well as a control group receiving no feedback were randomly assigned. The participants were pretested on the target structure of the study, the performance on present/past participle adjectives and showing no significant difference, they went under the treatment phase. The GA test, as the post test was performed after the treatment, and the gathered data were compared through the SPSS. Meanwhile a one-way ANOVA and Tukey's post-hoc tests were also used to compare the results. The criteria p< 0.05 was also considered for the group differences. The findings not only indicated the positive effect of CF, being either direct or indirect, but also the more effectiveness of direct CF on the learners' GA in SCE.