OBJECTIVES: The UKPDS 68 equations are routinely used to assess the benefits associated with glucose control in type 2 diabetes mellitus (T2DM) cohorts.  Their validity has been questioned because they were derived from a cohort of newly diagnosed T2DM subjects and diabetes management practices have evolved considerably. The objective of this study was to re-calibrate the equations to contemporary healthcare data and assess their continued relevance. METHODS: We utilized data from The Health Improvement Network (THIN) database over the period 1/1/2004 to 31/12/2009.  We selected patients either aged > 64 years with HbA1c >7.5% and diabetes duration >10 years or aged >55 years with at least 1 established cardiovascular risk factor. Weibull survival equations were fitted to the following endpoints; myocardial infarction (MI), stroke, congestive heart failure (CHF), ischemic heart disease (IHD), amputation, blindness and end-stage renal disease (ESRD) using R. Missing data was modeled using multiple imputation. RESULTS: Data were available on 68,990 T2DM subjects meeting the inclusion criteria with mean age 66.1 years, 46% female, 8.8 years duration of diabetes; with mean body mass index (29.7m/kg2), HbA1c (8.0%), systolic blood pressure  (147.5mmHg), total cholesterol (5.1mmol/l) and HDL cholesterol (1.3mmol/l).  Log hazards (standard error) associated with unit changes HbA1c were 0.097(0.011) for MI, 0.084 (0.008) for CHF, 0.051 (0.009) for stroke, 0.112(0.02) for blindness, 0.043 (0.007) for ESRD and 0.225 (0.017) for amputation.  These estimates were statistically consistent (at the 95% level) with the original UKPDS log hazards except for IHD (non-significant in the THIN database) and ESRD (non-significant in UKPDS). CONCLUSIONS: In general, the UKPDS equations retain their validity for assessing the relationship between HbA1c and macrovascular and microvascular complications.  It is likely that use of UKPDS equations will overestimate the incidence of IHD and under-estimate the incidence of ESRD.