Aim: To evaluate the degree of concordance between venous plasma glucose (PG) and capillary whole blood glucose (BG) in diagnosing Type 2 diabetes and to compare the prevalence of confirmed and unconfirmed Type 2 diabetes. Methods: All 40- to 69-year olds without known diabetes listed with five general practices were invited for screening (2051 persons). People identified by screening as being at high risk for having diabetes (random BG (rBG) ≥ 4.5 mmol/l or HbA1c ≥ 5.9%) were examined using BG and PG. The main outcome measures were confirmed diabetes cases (one diabetic glucose value plus symptoms or two diabetic glucose values on two different days) and unconfirmed diabetes (one diabetic glucose value on the first day of testing). Results: One thousand and twenty-eight (50%) accepted screening. Diabetes was confirmed by BG or PG in 22 patients (1.8%), and in 14 of these by both BG and PG. Four patients in whom diabetes was confirmed only by BG had unconfirmed diabetes by PG. Three of four people in whom diabetes was confirmed only by PG had unconfirmed diabetes according to BG. The prevalence of unconfirmed diabetes was 3.3% and the prevalence of confirmed diabetes 2.1%. Conclusion: Confirmed diabetes can be diagnosed by either BG or PG. BG seems more convenient than PG for use in general practice. The prevalence of confirmed diabetes was lower than the prevalence of unconfirmed diabetes.