Contract learning as an approach to individualizing education in the context of assessment for learning is relatively underexplored in English as a Foreign Language instruction. The present study used a mixed-methods design to investigate its efficacy to provide feedback to students and improve self-directed learning. Furthermore, it studied students' impression of the assessment procedure and their agency. After taking a diagnostic test of grammatical knowledge, participants (N = 14 graduate students) received treatment, using learning contracts, and then took a posttest. A paired t test showed a significant improvement in their grammatical knowledge (t = 7.96, df = 13, p = 0.00), the effect size being r = 0.91. The analysis of data from an open-ended questionnaire helped explain how the approach had a positive effect on learning of grammar. The results also indicated that the students had favorable attitudes toward contract learning, mentioning that it had a beneficial impact on their studying behavior. The article concludes with a discussion on the validity considerations of using contract learning in an assessment for learning context.