OBJECTIVE: To compare African-American patients (16% of the population) and Latino patients (15% of the population) to their respective complementary cohorts of all other patients (including 66% Caucasian, and 3% others) in terms of glycemic response to pioglitazone combination therapy. METHOD: Twenty-four-week data from three randomized controlled trials, each using 30- or 45mg doses of pioglitazone combined with i) sulfonylurea; ii) metformin; or iii) insulin for 24-weeks, were analyzed. Mean changes from baseline in hemoglobin A1c between (a) African-Americans and all others and (b) Latinos and all others were compared using regression models with baseline demographics and hemoglobin A1c as covariates. RESULTS: Baseline hemoglobin A1c measurements ranged from 10.11 to 10.86 among African-Americans, from 9.73 to 10.11 among Latinos, and from 9.57 to 9.89 among all others. The mean reduction in hemoglobin A1c from baseline was statistically significant in every African-American treatment group (-0.96% to –2.34%) and every Latino treatment group (-1.26% to -1.85%) as well as the “all others” groups. Significantly larger mean reductions in hemoglobin A1c were detected in African-Americans than all others in two treatment groups (receiving higher doses of pioglitazone with sulfonylurea and with metformin) and in Latinos in one treatment group (the lower dose of pioglitazone with metformin combination). Patients with relatively higher baseline hemoglobin A1c and body mass index measurements tended to benefit more from treatment. CONCLUSION: Data from three clinical trials show that African-Americans and Latinos attained decreases in hemoglobin A1c levels using pioglitazone combination therapy. Generally, the improvements in hemoglobin A1c in African-Americans and Latinos were comparable to the corresponding cohorts of all others.