Background: Thrombolysis in myocardial infarction risk score (TIMI-RS) was designed to predict early mortality in patients with a ST elevation acute myocardial infarction (STEAMI). Aim: To evaluate the predictive capacity for hospital mortality of TIMI-RS. Material and methods: Patients with ? 12-hour evolution STEAMI were selected from a prospective registry of all patients hospitalized in our coronary unity within January 1988 and December 2005. Observed mortality was analyzed according to TIMI-RS and its predictive capacity was estimated. Results: We analyzed 1125 consecutive patients aged 61±13 years (76% men). Fifty one percent were smokers, 47% hypertensive and 40% had a history of angina. Fifty eight percent of patients underwent reperfusion therapy. Most patients had TIMI-RS scores ?5 points and only 3.6% had scores ?10 points. Overall mortality was 14.8% and there was an 80% concordance between observed mortality and that predicted with the TIMI-RS score. The area under the curve for the receiver operating characteristic (ROC) curve was 0.7. Conclusions: TIMI-RS was acceptably useful to predict in-hospital mortality in this group of patients with STEAMI. Differences between the observed and originally predicted mortality are explained by the clinical profile and therapeutic protocols applied to patients in different studies. Thus, caution needs to be taken when interpreting the risk associated to a specific score, particularly within non-reperfused patients whose risk might be underestimated.Background: Thrombolysis in myocardial infarction risk score (TIMI-RS) was designed to predict early mortality in patients with a ST elevation acute myocardial infarction (STEAMI). Aim: To evaluate the predictive capacity for hospital mortality of TIMI-RS. Material and methods: Patients with ? 12-hour evolution STEAMI were selected from a prospective registry of all patients hospitalized in our coronary unity within January 1988 and December 2005. Observed mortality was analyzed according to TIMI-RS and its predictive capacity was estimated. Results: We analyzed 1125 consecutive patients aged 61±13 years (76% men). Fifty one percent were smokers, 47% hypertensive and 40% had a history of angina. Fifty eight percent of patients underwent reperfusion therapy. Most patients had TIMI-RS scores ?5 points and only 3.6% had scores ?10 points. Overall mortality was 14.8% and there was an 80% concordance between observed mortality and that predicted with the TIMI-RS score. The area under the curve for the receiver operating characteristic (ROC) curve was 0.7. Conclusions: TIMI-RS was acceptably useful to predict in-hospital mortality in this group of patients with STEAMI. Differences between the observed and originally predicted mortality are explained by the clinical profile and therapeutic protocols applied to patients in different studies. Thus, caution needs to be taken when interpreting the risk associated to a specific score, particularly within non-reperfused patients whose risk might be underestimated.