Following international guidelines, communities have deployed automated external defibrillators at train stations without substantive evidence. We geocoded public out-of-hospital cardiac arrests (OHCAs) (2016-2020), automated external defibrillators, and train stations. The stations were divided into the following groups according to passenger flow: 1 (0-499), 2 (500-4999), 3 (5000-9999), and 4 (>10 000) passengers per day. Risk ratios (RRs) were calculated using Poisson regression of rates, and odds ratios (ORs) were analyzed through logistic regression. OHCAs at train stations accounted for 102 (2.3%) of 4467 public OHCAs. The incidence rate (IR) and RR for OHCAs were for group 1: IR, 0.02 OHCA per station per year, RR, 1.0 (reference); group 2: IR, 0.07, RR, 4.1 (95% CI, 2.3-7.3); group 3: IR, 0.25, RR, 12.7 (95% CI, 6.2-25.9); and group 4: IR, 0.34, RR, 16.3 (95% CI, 8.6-30.9). Compared with other public OHCAs, OHCAs at train stations were just as likely to receive bystander cardiopulmonary resuscitation (OR, 1.13 [95% CI, 0.60-2.12]). However, they had higher odds of bystander defibrillation (OR, 1.66 [95% CI, 1.06-2.58]), were more likely to achieve return of spontaneous circulation (OR, 1.88 [95% CI, 1.24-2.85]), and survive 30 days (OR, 2.37 [95% CI, 1.57-3.59]). The incidence of OHCAs at train stations was associated with passenger flow, with the busiest stations having a 16-fold higher risk of OHCAs than the lowest. OHCAs at train stations had better outcomes compared with other public OHCAs.