This article reports on an investigation of the impact of different methods of journaling on self-efficacy of learners of English as a foreign language. Sixty upper-intermediate Iranian English language learners were randomly assigned to three experimental conditions, namely no-feedback, teacher-feedback, and peer-feedback, and one control group--no-journal condition. The no-feedback condition learners kept reflective journals but could not share their reflective notes with other learners or their teacher, while the teacher-feedback condition learners used collaborative reflective journals with their teacher. The peer-feedback condition learners shared collaborative reflective journals with their peers, and the no-journal condition learners did not use reflective journals throughout the treatment period. An analysis of variance showed that the students with the two feedback conditions seemed to have gained a higher self-efficacy at the end of the treatment period.