The present study examined whether a web-based intervention could promote weight-loss maintenance, after weight loss. The study was a two-phase, 27-month, randomized controlled trial conducted in Ibaraki, Japan, from 2014 to 2017; 133 participants were recruited through local newspaper advertisements. The eligibility criteria were as follows: age of 40-64 years, body mass index of 25-40 kg/m, and having at least one metabolic syndrome component. In phase 1, a 3-month, group-based weight-loss program was provided to all eligible participants ( = 119). We then randomly assigned (1:1) participants who had lost 5% or more of their weight during phase 1 ( = 95) to either the self-help (mean 3-month weight loss 7.30 kg) or the web-support group (7.00 kg). Participants in the web-support group regularly reported their body weight and physical activity through a web-based system. They received monthly personalized feedback from a study staff for 24 months. The primary outcome, 27-month body-weight change (mean ± standard deviation), in the self-help and web-support groups were - 5.3 ± 5.0 kg and -4.5 ± 4.9 kg, respectively. There was no significant difference. An exploratory secondary analysis demonstrated that those with greater 27-month increases in their step count, assessed with an accelerometer, lost more weight with no difference in changes in energy intake. The mean 27-month body-weight change in the 4th quartile of changes in step count was -7.78 kg. Although web-based intervention using an activity monitor failed to promote weight-loss maintenance, increased physical activity was associated with successful weight-loss maintenance.