This study investigated the effects of recast and explicit corrective feedback on Iranian IELTS test-takers anxiety in speaking across different levels of proficiency. Ninety male and female learners, aged 21 to 45, who were preparing themselves to take the IELTS Mock test, were divided into upper-intermediate and advanced levels. Then, they were randomly assigned to control and experimental groups (recast and explicit correction). There were 10 treatment sessions for each group. In the recast group, the students’ mistakes were corrected using reformulation. In the explicit correction group, mistakes were not tolerated, and they were corrected on the spot. In order to measure the anxiety in speaking performance, Chowdhury’s (2014) questionnaire was given to the participants once before the treatment and once after the treatment. Then, the score of each learner was calculated and recorded for the data analysis. The results showed that there was no statistically significant difference between the control and explicit corrective group and the control and recast groups at the upper-intermediate level. However, there was a statistically significant difference between the control and recast groups and the control and explicit group at the advanced levels.
