We studied the efficacy of four different treatment regimens (sulphonylurea and metformin+/-acarbose versus glimepiride and rosiglitazone versus glimepiride and bedtime NPH insulin versus multiple actrapid and NPH insulin injections) in poorly controlled type 2 diabetes subjects on hs-CRP, VCAM-1 and AGE at 4, 8 and 12 weeks of treatment. Multiple insulin injections rapidly improved HbA(1c) by 0.6+/-0.9% (p<0.005), 1.2+/-1.3% (p<0.0005) and 1.3+/-1.4% (p<0.0005) at week 4, at week 8 and week 12, respectively. Subjects who continued their existing combination treatment of sulphonylurea, metformin+/-acarbose also showed a significant reduction in HbA(1c) (p<0.05). Although effective in reducing glycemic parameters, there was no reduction in CRP levels in either treatment group. The treatment regimen consisting of rosiglitazone and glimepiride significantly lowered hs-CRP by -2.6 (3.9) mg/L (p<0.05) at week 12 in spite of no improvement in blood glucose. AGE improved in all groups irrespective of type of treatment, glycaemic control and CRP levels. Our data indicate rapid glycaemic control alone does not necessarily result in improvement in markers of inflammation in type 2 diabetes patients.