Aims: This pilot study aimed to assess the acceptability of an intensive community-based intervention, and to get a preliminary estimate of the effect size of the intervention for people with type 2 diabetes in Kupang, Indonesia. Methods: Pilot of a cluster randomised trial involving sixty-seven participants. The participants were recruited from four Primary Health Centres (PHCs), and were allocated into groups (intervention: 36; control: 31) based on the PHCs they were registered in. The 3 month intervention consisted of weekly group activities including physical activity and education. Control was usual care. The outcomes were change in HbA1c, medication adherence, and self-management behaviours. To test our intervention, we used a mixed linear model, with participants nested within cluster/health centre. Results: Participants were 60% female, mean age 57, and had diabetes for an average of 4.5 years. Groups did not differ at baseline. Counter to prediction, no effects of the intervention over usual care were observed on the outcomes measured. However, both groups made statistically significant improvements in HbA1c (intervention from 8.3% +/- 1.9 to 7.8% +/- 1.8; control from 8.4% +/- 2.6 to 7.9% +/- 2.5). Conclusions: The effect of the intervention did not differ from usual care. However, both groups made improvements in HbA1c outcome, suggesting that the project might have indirectly triggered more intensive care for diabetes patients. (c) 2021 Elsevier B.V. All rights reserved.