Background The widespread adoption of high-sensitivity cardiac troponin testing has encouraged the use of pathways to accelerate the rule-out and rule-in myocardial infarction in the Emergency Department. These pathways are not recommended for patients with ST-segment elevation, but there is a risk they may be applied incorrectly given that interpretation of the electrocardiogram is subjective, dependent on experience, and signs may be masked in those with posterior myocardial infarction. Methods Consecutive patients with suspected acute coronary syndrome were enrolled in a stepped-wedge cluster randomized controlled trial across ten hospitals in Scotland. The index diagnosis was adjudicated two clinicians independently in all patients with high-sensitivity cardiac troponin I concentrations above the sex-specific 99th centile on serial testing and abnormalities on the electrocardiogram recorded. The proportion of patients with ST-segment elevation myocardial infarction and concentrations below the rule-out threshold (<5 ng/L), 99th centile (<16 ng/L and <34 ng/L for women and men) and rule-in threshold (<52 ng/L) at presentation were determined. Secondary analysis determined the effect of symptom duration, and culprit vessel location, on troponin concentrations. Results In total, we enrolled 48,282 consecutive patients with suspected myocardial infarction were enrolled, with 925 having an index diagnosis of STEMI. The majority (83.5%, 772/925) of patients had a troponin concentration above the 99th-centile on presentation. The median troponin concentration on presentation was 196 ng/L [46.0, 21611.], with 2.2% (20/925) and 14.4% (133/925) under the 5ng/L rule-out threshold, and <99th-centile respectively. Relying on a rule-in threshold of 52ng/L would miss more than 1 in 4 patients (26.8%, 248/925) with STEMI. Patients presenting soon after symptom onset had significantly lower troponin concentrations (<2 hours, median 96.0ng/L [26.0, 494.0] vs ≥2 hours, median 294.ng/L [59.0, 3042.0], p<0.001).