Aim: To examine the preliminary effects of a culturally tailored, family-supported, community-based diabetes self management education and support (DSMES) programme for Ethiopian people with type 2 diabetes on glycosylated haemoglobulin (HbA(1c)), blood pressure, body mass index and lipid profiles. Methods: A two-arm pilot randomised controlled trial (RCT) was conducted involving 76 participant-caregiver dyads from Western Ethiopia, which were randomly allocated to the intervention arm to receive 12 h of DSMES intervention guided by social cognitive theory on top of usual care, or to the control group, which received usual care. While HbA(1c) was a primary outcome, the blood pressure, body mass index and lipid profiles were secondary outcomes. Primary outcome was the change in HbA(1c) between baseline and 2-month follow-up between the groups. Generalised estimating equations was used to test the preliminary effect of the DSMES programme on the outcomes at baseline, post-intervention and at 2-month follow-up for secondary outcomes. Cohen's d was used to estimate the between-group effect sizes of the intervention. Results: The DSMES produced significant improvement in HbA(1c) with large effect size (beta = -1.667, p < 0.001, d = -0.81) and triglycerides with medium effect size (d = -0.50). HbA(1c) in the intervention group was decreased by 12 mmol/mol (1.1%). Although nonsignificant, the DSMES also had small to moderate effects (d = -0.123 to 0.34) on blood pressure, body mass index, total cholesterol, low-density and high-density lipoproteins when compared with usual care. Conclusion: A culturally tailored, social cognitive theory-guided, family-supported, community-based DSME programme could have a benefit on HbA(1c) and triglycerides. A full RCT is warranted to test the effectiveness of the DSMES programme.