OBJECTIVE: Cut-offs for high-sensitivity troponin (hs-Tn) elevations to define prognostically significant peri-operative myocardial injury (PMI) in cardiac surgery is not well-established. We evaluated the associations between peri-operative high-sensitivity troponin T (hs-TnT) elevations and 1-year all-cause mortality in patients undergoing cardiac surgery. METHODS: The prognostic significance of baseline hs-TnT and various thresholds for post-operative hs-TnT elevation at different time-points on 1-year all-cause mortality following cardiac surgery were assessed after adjusting for baseline hs-TnT and EuroSCORE in a post-hoc analysis of the ERICCA trial. RESULTS: 1206 patients met the inclusion criteria. Baseline elevation in hs-TnT >x1 99th percentile upper reference limit (URL) was significantly associated with 1-year all-cause mortality (adjusted hazard ratio 1.90, 95% confidence interval 1.15-3.13). In the subgroup with normal baseline hs-TnT (n = 517), elevation in hs-TnT at all post-operative time points was associated with higher 1-year mortality, reaching statistical significance for elevations above: ≥100 × URL at 6 h; ≥50 × URL at 12 and 24 h; ≥35 × URL at 48 h; and ≥30 × URL at 72 h post-surgery. Elevation in hs-TnT at 24 h ≥ 50 × URL had the optimal sensitivity and specificity (73% and 75% respectively). When the whole cohort of patients was analysed, including those with abnormal baseline hs-TnT (up to 10 × URL), the same threshold had optimal sensitivity and specificity (66% and 70%). CONCLUSIONS: Both baseline and post-operative hs-TnT elevations are independently associated with 1-year all-cause mortality in patients undergoing cardiac surgery. The optimal threshold to define a prognostically significant PMI in our study was ≥50 × URL elevation in hs-TnT at 24 h.