To examine prospectively the relationship between diabetes, glycemic control, and spirometric measures. From a community-based cohort, 495 Europid (i.e., of European descent) patients with type 2 diabetes who had no history of pulmonary disease underwent baseline spirometry between 1993 and 1994. A subset of 125 patients was restudied a mean of 7.0 years later. The main outcome measures included forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), vital capacity (VC), and peak expiratory flow (PEF) corrected for body temperature, air pressure, and water saturation and were expressed either in absolute terms or as percentage-predicted value for age, sex, and height. Mean percentage-predicted values of each spirometric measure were decreased >10% in the whole cohort at baseline and absolute measures continued to decline at an annual rate of 68, 71, and 84 ml/year and 17 l/min for FVC, FEV1, VC, and PEF, respectively, in the 125 prospectively studied patients. Declining lung function measures were consistently predicted by poor glycemic control in the form of a higher updated mean HbA1c, follow-up HbA1c, or follow-up fasting plasma glucose. In a Cox proportional hazards model, decreased FEV1 percentage-predicted value was an independent predictor of all-cause mortality. Reduced lung volumes and airflow limitation are likely to be chronic complications of type 2 diabetes, the severity of which relates to glycemic exposure. Airflow limitation is a predictor of death in type 2 diabetes after adjusting for other recognized risk factors.