The Predictable Results and Experience in Diabetes through Intensification and Control to Target: An International Variability Evaluation 303 (PREDICTIVE 303) Study (n = 5604) evaluated the effectiveness of insulin detemir, a long-acting basal insulin analogue, using a simplified patient self-adjusted dosing algorithm (303 Algorithm group) compared with standard-of-care physician-driven adjustments (Standard-of-care group) in a predominantly primary care setting, over a period of 6 months. Insulin detemir was to be started once-daily as add-on therapy to any other glucose-lowering regimens or as a replacement of prestudy basal insulin in patients with type 2 diabetes. Investigator sites rather than individual patients were randomized to either the 303 Algorithm group or the Standard-of-care group. Patients from the 303 Algorithm group sites were instructed to adjust their insulin detemir dose every 3 days based on the mean of three 'adjusted' fasting plasma glucose (aFPG) values (capillary blood glucose calibrated to equivalent plasma glucose values) using a simple algorithm: mean aFPG < 80 mg/dl (<4.4 mmol/l), reduce dose by 3 U; aFPG between 80 and 110 mg/dl (4.4-6.1 mmol/l), no change; and aFPG > 110 mg/dl (>1.1 mmol/l), increase dose by 3 U. The insulin detemir dose for patients in the Standard-of-care group was adjusted by the investigator according to the standard of care. Mean A1C decreased from 8.5% at baseline to 7.9% at 26 weeks for the 303 Algorithm group and from 8.5 to 8.0% for the Standard-of-care group (p = 0.0106 for difference in A1C reduction between the two groups). Mean FPG values decreased from 175 mg/dl (9.7 mmol/l) at baseline to 141 mg/dl (7.8 mmol/l) for the 303 Algorithm group and decreased from 174 mg/dl (9.7 mmol/l) to 152 mg/dl (8.4 mmol/l) for the Standard-of-care group (p < 0.0001 for difference in FPG reduction between the two groups). Mean body weight remained the same at 26 weeks in both groups (change from baseline 0.1 and -0.2 kg for the 303 Algorithm group and the Standard-of-care group respectively). At 26 weeks, 91% of the patients in the 303 Algorithm group and 85% of the patients in the Standard-of-care group remained on once-daily insulin detemir administration. The rates of overall hypoglycaemia (events/patient/year) decreased significantly from baseline in both groups [from 9.05 to 6.44 for the 303 Algorithm group (p = 0.0039) and from 9.53 to 4.95 for the Standard-of-care group (p < 0.0001)]. Major hypoglycaemic events were rare in both groups (0.26 events/patient/year for the 303 Algorithm group and 0.20 events/patient/year for the Standard-of-care group; p = 0.2395). In conclusion, patients in the 303 Algorithm group achieved comparable glycaemic control with higher rate of hypoglycaemia as compared with patients in the Standard-of-care group, possibly because of more aggressive insulin dose adjustments. The vast majority of the patients in both groups were effectively treated with once-daily insulin detemir therapy. The use of insulin detemir in this predominantly primary care setting achieved significant improvements in glycaemic control with minimal risk of hypoglycaemia and no weight gain.