OBJECTIVE: Coronary artery disease (CAD) is a leading cause of mortality and morbidity in diabetic patients; therefore, their risk stratification is a relevant issue. Because exercise tolerance is frequently impaired in these patients, pharmacological stress echocardiography (SE) has been suggested as a valuable alternative. Our aim was to evaluate the prognostic value of this technique in diabetic patients with known or suspected CAD. RESEARCH DESIGN AND METHODS: A total of 259 consecutive diabetic patients underwent pharmacological SE (dobutamine in 108 patients and dipyridamole in 151 patients) and follow-up for 24 +/- 22 months. A comparison between the prognostic value of SE and exercise electrocardiography (ECG) was made in a subgroup of 120 subjects. RESULTS: A total of 13 cardiac deaths and 13 nonfatal infarctions occurred during follow-up, and 58 patients were revascularized. Univariate predictors of outcome were known CAD, positive SE, rest and peak wall motion score index (WMSI), and peak/rest WMSI variation. Peak WMSI was the only significant and independent prognostic indicator (odds ratio 11; 95% CI 4-29, P < 0.0001) on multivariate Cox's analysis. After adjustment for the most predictive clinical and exercise ECG variables, SE provided 43% additional prognostic information (gain in X(2) = 7, P < 0.01). Moreover, positive SE was associated with a significantly lower event-free survival. CONCLUSIONS: SE effectively predicts cardiac events in diabetic patients with known or suspected CAD and adds additional prognostic information as compared with exercise ECG.