OBJECTIVE: To determine the effect of thiazolidinedione treatment on subclinical atherosclerosis progression in insulin-requiring patients with clinical characteristics suggesting type 2 diabetes. RESEARCH DESIGN AND METHODS: Eligible participants (n = 299) were randomized within strata of baseline common carotid artery (CCA) intima-media thickness (IMT) (<0.8 mm, >or=0.8 mm) to 400 mg troglitazone daily or placebo for 2 years. A general linear mixed-effects model was used to compare the rate of change in CCA-IMT between treatment groups. RESULTS: Overall, average rates of CCA-IMT change were not significantly different between troglitazone- and placebo-treated subjects (0.0030 +/- 0.021 vs. 0.0066 +/- 0.021 mm/year; P = 0.17). In the stratum of subjects with CCA-IMT >or=0.8 mm, troglitazone significantly reduced the progression of CCA-IMT relative to placebo (0.0013 +/- 0.022 vs. 0.0084 +/- 0.023 mm/year; P = 0.03). Fasting glucose, insulin, and HbA(1c) were significantly lower in troglitazone- versus placebo-treated subjects (P < 0.01). Whereas blood pressure significantly differed between treatment groups in the >or=0.8-mm stratum, there was no difference between treatment groups in the <0.8-mm stratum. CONCLUSIONS: Insulin sensitization and reduction in blood pressure may be contributory mechanisms by which troglitazone reduced subclinical atherosclerosis progression in this cohort of well-controlled insulin-dependent patients with clinical characteristics suggesting type 2 diabetes.