OBJECTIVES: The ongoing Cardiovascular (CV) Risk Evaluation in people with Type-2 diabetes mellitus (T2DM) on Insulin Therapy (CREDIT) study is assessing the effect of insulin on the risk of vascular events. METHODS: CREDIT is a 4-year, 314 centre, non-interventional trial in North America, Europe and Asia, and includes 3031 people with T2DM who recently started basal and/or short-acting insulin, premix insulin or another insulin type. This analysis examines and compares the characteristics between groups starting basal (n=1563), basal+short-acting (n=444), short-acting (n=221), premixed (n=700) or another (n=103) insulin. RESULTS: Demographic and diabetes characteristics were reasonably balanced between the insulin groups, although those receiving basal plus short-acting insulin or premix had a trend to higher baseline HbA1c levels vs other insulin types (basal, 9.2±1.8%; basal+short-acting, 10.1±2.2%; short-acting, 9.4±2.0%; premix, 9.9±2.0%; other, 9.1±2.0%). While the majority had previously used oral glucose lowering drugs (OGLDs) (basal, 97%; basal+short-acting, 83%; short-acting, 83%; premix, 94%; other, 85%), differences in the numbers continuing OGLDs when beginning insulin were found. Continued use of OGLDs was highest wth basal insulin (89%) versus the other insulins (basal+short-acting, 36%; short-acting, 45%; premix, 62%; other, 34%). However, the distribution of types of OGLD used before insulin was similar between the groups. There are no clear patterns in CV risk profile by insulin type. Previous diagnosis of hypertension (basal, 71%; basal+short-acting, 65%; short-acting, 57%; premix, 69%; other, 72%), family history of CV disease (basal, 29%; basal+short-acting, 25%; short-acting, 21%; premix, 23%; other, 14%) and body mass index tended to be lower in the short-acting insulin group. However, triglyceride levels were lower in the short-acting and 'other' insulin groups vs premix, basal and basal plus short-acting groups. CONCLUSIONS: People starting different insulins have somewhat different clinical characteristics, which may confound attempts to compare future vascular outcomes between regimens.