yThere has been a prolonged argument on the usefulness of the corrective feedback on the accuracy and fluency of students’ writings since Truscott's (1996) declaration that grammar correction had harmful effects on ESL/EFL learners' writing ability. The main purpose of this study was to investigate and compare the possible effects of direct and indirect on accuracy and fluency in subsequent and revision compositions. The partakers were selected; the investigation took place on 38 grade 12 students at Amanuel Preparatory School. The students were grouped into three randomly; they were received direct, indirect and no feedback conditions. Six consecutive writing tasks were equally given for each group for 9 weeks. In accuracy, error ratios (the number of errors counted and divided by the number of word production times hundred) were calculated; fluency equal to time spent to write a text divided by word production times hundred. It was carried out by likening the effects of direct and indirect together with control group through one way ANOVA. The numerical results were strengthened by textual analysis. The general findings revealed that direct and indirect had equal significant effects in subsequent tasks, whereas on the revision essays although both WCFs had significant effects, direct was a little superior to indirect. Another result of this study showed that neither DWF (direct) nor IDWF (indirect) had possible significant effects on subsequent and revision compositions on fluency. The specific findings depicted that out of ten specified categories of error, the seven types of error in the students compositions were capitalization (31.2%), spelling (19.28%), punctuation (18.6.82%), tense (11.25%), missed word (8.79%), verb form (4.22%), and number (3.25%); more specifically, DWF was finer to IDWF to reduce capitalization, spelling and verb form errors in the subsequent essays; and also it was superior to IDWF for the reduction of spelling, missed word, tense and verb form errors in revision essays, while IDWF was better than DWF for decreasing punctuation errors in both cases and missed word errors on new essays. On the other hand, other findings depicted that both DWF and IDWF were equally effective for lowering capitalization in revision and tense errors in subsequent essays. But DWF and IDWF had not significant variation for decreasing number errors with NF group; exceptionally, IDWF was ineffective for reducing spelling errors in both cases. The textual analysis directly supported the above findings qualitatively. The output revealed that students in the experimental groups decreased their errors far significantly than students in the NF group. Throughout the study, there was no significant difference between DWF and IDWF. So, providing both WCFs on students ‘compositions should be essential technique to improve accuracy without hindering fluency.