High serum alkaline phosphatase (ALP) levels are associated with excess all-cause and cardiovascular mortality in patients undergoing hemodialysis (HD). However, the long-term relationship between serum ALP levels and infection-related mortality remains unclear. A total of 3502 maintenance HD patients were registered in the Q-Cohort Study, an observational cohort study in Japan. The primary outcome was infection-related mortality during a 10-year follow-up period. The covariate of interest was serum ALP levels at baseline. The association between serum ALP levels and infection-related mortality was calculated using a Cox proportional hazards model and a Fine-Gray subdistribution hazards model with non-infection-related death as a competing risk. During the follow-up period, 446 patients died of infection. According to their baseline serum ALP levels, the patients were categorized into sex-specific quartiles (Q1-Q4). Compared with patients in the lowest serum ALP quartile (Q1), those in the highest quartile (Q4) had a significantly higher multivariable-adjusted hazard ratio (HR) of 1.70 [95% confidence interval (CI) 1.24-2.32] for infection-related mortality. Furthermore, the HR for every 50U/L increase in serum ALP levels was 1.24 (95% CI 1.12-1.36) for infection-related mortality. These associations remained consistent in the competing risk model: subdistribution HR, 1.47; 95% CI 1.07-2.03 for Q4 compared with Q1. Higher serum ALP levels were significantly associated with a higher risk of infection-related mortality in patients undergoing HD.