OBJECTIVES: The American Diabetes Association recommends treating diabetes patients to a target hemoglobin A1c level of 7% or less. Our objective was to examine the influence of A1c test results on decisions to initiate exogenous insulin supplementation in Type 2 diabetes. METHODS: A database of clinical laboratory results linked to health plan claims (MarketScan Lab) was used to identify patients with a diagnosis of Type 2 diabetes. Their first A1c test result between July 2003 and June 2005 was identified and patients were classified as at target (less than 7), above target (7-9), or poorly controlled (>9). The study sample was restricted to patients who had 6 months of claims history prior to and 6 months after the index test, and who received only oral antidiabetes therapy during the pre-test period. Addition of insulin to the regimen within 6 months after the test was assessed. RESULTS: A total of 4,836 patients were identified, including 2,450 (51%) who were at target, 1,567 (32%) above target, and 819 (17%) poorly controlled. Insulin was added to the regimen of 12.9% of patients with poor control, 4.3% of those above target, and 1.5% of those at target (p<0.00001). Among poorly controlled patients insulin was added in 22.0% of patients already on 3 oral medications, 12.6% of those with 2 oral medications, and 8.6% of those with 1 oral medication (p<0.001). A1c test result and prior oral medication use remained significant predictors of insulin supplementation after controlling for demographic variation. CONCLUSION: High levels of A1c indicating poor glycemic control led some patients with Type 2 diabetes to initiate insulin supplementation, particularly those already receiving multiple oral medications. However, poorly controlled patients were still most likely to remain on oral therapy only, even if they were already using 3 or more medications.