The objective of the study was to investigate the association between alkaline phosphatase (AP) activity and prognosis of patients with acute coronary syndrome (ACS).The study included 2134 patients with ACS undergoing percutaneous coronary intervention. All included patients had baseline AP measurements available. The receiver operating characteristic curve analysis showed that the best cut-off of AP for mortality prediction was 98.0U/L. Using this cut-off, patients were divided into two groups: a group with AP>98.0U/L (n=493) and a group with AP<=98.0U/L (n=1641). The primary endpoint was 3-year mortality.Overall, there were 229 deaths over the follow-up: 90 deaths among patients with an AP >98.0U/L and 139 deaths among patients with an AP<=98.0U/L (Kaplan-Meier estimates of 3-year total mortality, 19.5% and 9.3%, respectively; adjusted hazard ratio [HR]=1.37, 95% confidence interval [CI] 1.10-1.70, P=0.004 for each unit higher log AP). Cardiac deaths occurred in 157 patients: 66 deaths among patients with an AP>98.0U/L and 91 deaths among patients with an AP<=98.0U/L (Kaplan-Meier estimates of 3-year cardiac mortality, 14.3% and 6.0%, respectively; adjusted HR=1.32 [1.02-1.70], P=0.033, for each unit higher log AP). The C-statistic of the multivariable model with baseline variables was 0.836 [0.807-0.866] and it increased to 0.842 [0.814-0.874] after inclusion of AP (P=0.045).In patients presenting with an ACS and treated with percutaneous coronary intervention, elevated AP activity is associated with increased risk of subsequent mortality.