Background: Patients with chest pain represent a large portion of emergency department (ED) presentations around the world. Accurate risk assessment to rule out acute coronary syndrome (ACS) is required. Aim: To measure accuracy and characteristics of ED doctors and nurses when assessing patients for suspected ACS and the associations with outcome. Methods: A sub-analysis of a large randomised trial that compared high-sensitivity troponin to conventional troponin was carried out. Both nurses and doctors conducted ACS risk assessments on 1857 patients and concordance comparisons were undertaken. Findings: There was no difference between the accuracy of ED nurses and doctors (c-statistic: 0.67 vs 0.68 respectively; P=0.35). Both clinician types made similar ACS risk assessments. Nurses with specialist qualifications and consultant doctors were the most accurate, at 65.4% and 74.2%, respectively. Conclusion: Nurses and doctors in the ED made similar ACS risk assessments but both were sub-optimal. Standard clinical pathways to assist ED doctors and nurses improve their ACS impression accuracy could be beneficial.