Purpose: The benefits of teacher autonomy support for optimal school functioning is evident in literature. However, studies are meager about teachers applying the concept of autonomy support in real settings (regular classroom). Design/methodology/approach: Based on empirical data within self-determination theory, a longitudinal intervention program was designed to facilitate autonomy support instruction in a natural classroom setting and to assess its effectiveness on Thai students' learning motivation. In a quasi-experimental, nonequivalent control group design, with an appended withdrawal design, 103 students from Grade 6 of a Thai public school underwent the intervention for seven weeks in natural classroom settings. The students in both the groups self-reported their class-related experience before intervention (pretest), after intervention (posttest1) and after the withdrawal of intervention (posttest2). Findings: MANOVA results revealed a significant mean difference for all dependent measures on posttest1 between the experimental group and control group. Also, after the withdrawal of treatment, the experimental group showed a significant omnibus effect on combination of dependent measures, while scores of control group were stable. Originality/value: The findings have implications for elementary school teachers in Thailand to adapt and adopt autonomy supportive instruction as a classroom practice.