Background. Child obesity is an increasing trend in Thailand. Increasing physical activity is therefore necessary. Objectives. The purposes of this study were to investigate the effect of a walking intervention program on the self-efficacy, physical activity (PA), and physical fitness of obese children in Thailand and to examine whether self-efficacy mediated changes in PA and physical fitness. Methods. A total of 42 children aged 6-19 years were randomized into two parallel groups: a control group (n=21) and an intervention group (n=21) who participated in the walking physical activity enhancement program for 8 weeks. Self-efficacy, %body fat, cardiovascular endurance, sedentary behavior (SB), steps, metabolic equivalents (METs), light PA (LPA), and moderate to vigorous PA (MVPA) were measured at baseline and after the end of the intervention. Multivariate analysis of variance tests was performed. METs and steps scores were highly positive; a composite score was created (i.e., Mets-Steps). Path analysis was conducted for mediation testing. Results. Compared to the control group, the intervention group had a significant increase in self-efficacy (p<0.05), Mets-Steps (p<0.01), and the MVPA (p<0.01) at the end of the program, while % body fat (p=0.41), cardiovascular endurance (p=0.11), SB (p=0.22), and LPA (p=0.31) were unaffected. Self-efficacy did not significantly mediate the effect of the intervention on METs-Steps and MVPA. Conclusion. The intervention can increase self-efficacy, steps, METs, and MVPA, but the increase in steps, METs, and MVPA is not a result of the increase in self-efficacy.