We conducted a randomized controlled trial for 12 weeks in patients with type 2 diabetes living in a rural part of Korea. The intervention group (n = 35) was managed by a diabetes centre which provided specialized management mediated by a primary health-care nurse who used a PDA-type blood glucometer with a bar code detector to measure the capillary glucose levels. The control group (n = 36) received usual care. Compared with baseline, HbA(1c) was significantly reduced at three-month follow-up in the intervention group (8.0% vs. 7.5%; P < 0.01), but not in the control group. Total cholesterol was significantly reduced in the intervention group (10.7 mmol/L vs. 10.4 mmol/L; P = 0.043). Fasting plasma glucose and triglyceride levels were lower at follow-up in both groups, but the difference was not significant. The new system could be implemented widely and would contribute to improving the quality of diabetes care, even in isolated rural areas.