OBJECTIVE: This study compared glycemic control in finger tip versus forearm sampling methods of self-monitoring of blood glucose (SMBG). RESEARCH DESIGN AND METHODS: One hundred seventy-four insulin-using patients with type 2 diabetes were randomized to SMBG using either finger-tip testing (FT) or forearm alternative site testing (AST) and followed up for 7 months. Hemoglobin A1C (HbA1C) was measured at baseline, month 4, and month 7. The study was designed to test the noninferiority of the AST method for the primary end point of change in HbA1C from baseline to month 7. Adherence with the testing schedule and frequency of hypoglycemic episodes were also measured. RESULTS: The FT (n = 85) and AST (n = 89) groups each had significant decreases in mean HbA1C from baseline to month 7 (FT, -0.4 +/- 1.4%, P = 0.008; AST, -0.3 +/- 1.2%, P = 0.045), and noninferiority between groups was demonstrated with a margin of equivalence of 0.5 (P = 0.043). There was no observable difference in HbA1C change between the groups (P = 0.442). Adherence was better in the FT (87%) than the AST (78%) group (P = 0.003), which may have been because of the difficulty some subjects had in obtaining blood samples for AST. The number of hypoglycemic episodes was too small to assess for a difference between groups. CONCLUSIONS: SMBG by the AST, rather than FT, method did not have a detrimental effect on long-term glycemic control in insulin-using patients with type 2 diabetes. Although adherence with testing was expected to be better in the AST group, it was actually better in the FT group.