Background: Acute coronary syndrome (ACS) is the first cause of mortality in Colombia. An erroneous risk stratification, in the emergency room (ER), affects the interventions performed and the rate of major cardiovascular adverse events. We measured the difference in GRACE score and stratification of coronary risk, by using the results of troponins measured sequentially during initial care. Methods: With a retrospective descriptive design, clinical records of patients treated for precordial pain of ≥ intermediate probability for ACS were evaluated, without indication of immediate invasive management, attended in the ER of a clinic of the third level of Bogotá, during 2017. Determined the difference between the GRACE score calculated with the first (GRACE-1), second (GRACE-2) or troponin delta (GRACE-delta) [paired T-test], and the proportion of poorly stratified patients was measured when using the first troponin [X2, Z-score]. Results: 44 patients in a period of 6 months were identified. The majority men, older adults, middle age 73 years. The average (SD) of scores GRACE-1, GRACE-2 and GRACE-delta, was 114.14 (30.73), 115.55 (30.14) and 111.11 (28.79), respectively; when comparing GRACE-delta with GRACE-1 and GRACE-2 significant differences were identified (p: <0.05). Error in the stratification of coronary risk was identified in 10/44 patients (22.7%), and 9/44 (20.4%) presented over-stratification. Conclusion: The stratification of coronary risk using the first troponin, unlike the troponin delta (item not clarified in the guidelines), evidenced an over-stratification in at least 20% of the patients, establishing the need for more invasive procedures and possibly longer hospital stay.