The extent of interaction between smoking and diabetes has been under-appreciated. Smokers had more diabetes, and when diabetes patients smoke, the combined mortality effect was greater than either the addition or multiplication of these two medical problems. Patients seen in the office are usually more interested in reducing blood glucose than in quitting smoking, and yet, smoking caused mortality risks, at a magnitude similar to or more than diabetes. The concept of "glucose equivalent of smoking" was developed to direct more attention to smoking in clinical management. Based on the follow-up observations from a large Asian cohort, the risk of an individual who smokes, from all-cause mortality, was found to be equivalent to an elevation of blood glucose by an average of 41 mg/dl for the cohort in general and 68 mg/dl for the diabetes in particular. By relating the message of smoking hazards in terms of "glucose equivalent", clinicians will be more alerted to counsel and patients will be more likely to quit. Appreciating this concept has a potential to change the paradigm of diabetes management, to bridge the clinical disconnect between the two, and to provide new ammunition for the diabetes epidemic in Taiwan. (c) 2005 Elsevier Ireland Ltd. All rights reserved.