Objective: This post-hoc analysis described efficacy and safety of insulin lispro low mixture (LM) twice daily versus bedtime insulin glargine plus prandial insulin lispro administered once daily before the largest meal (IGL) in East Asian (EA) and Caucasian type 2 diabetes mellitus patients who previously failed to reach glycemic targets on basal insulin glargine with metformin and/or pioglitazone. Methods: This phase 4, randomized, open-label, multinational, multicenter trial included patients with glycosylated hemoglobin (HbA1c) between ≥ 7.5% and ≤ 10.5%, and fasting plasma glucose ≤ 6.7 mmol/L. Results: Baseline mean (SD) HbA1c values were numerically similar between groups in EA (N = 79) and Caucasian (N = 278) patients (EA: LM = 8.78 [0.71] vs. IGL = 8.72 [0.90]; Caucasian: LM = 8.56 [0.78] vs. IGL = 8.51 [0.70]). Mean (SD) HbA1c significantly (P < 0.001) decreased from baseline to 24 weeks for both treatments in both subpopulations (EA: LM = -1.32% [0.96] and IGL = -0.89% [0.98]; Caucasian: LM = -1.24% [0.98] and IGL = -1.05% [0.97]). The proportion reaching HbA1c ≤ 7% at week 24 was LM = 33.3% and IGL = 22.9% for EA patients, and LM = 37.2% and IGL = 34.1% for Caucasian patients. Mean (SD) rate of hypoglycemia per 30 days was LM = 0.74 (1.16) and IGL = 1.22 (1.36) in EA patients, and LM = 1.38 (2.04) and IGL = 1.65 (2.43) in Caucasian patients. Mean (SD) change in weight gain was LM = 0.62 kg (2.78) and IGL = 0.51 kg (2.63) in EA patients, and LM = 1.77 kg (2.91) and IG = 0.67 kg (3.09) in Caucasian patients. Conclusion: Improved glycemic control was observed in EA and Caucasian patients treated with either LM or IGL.