This paper compares urban rail cost–benefit analysis (CBA) in 11 countries and the region of Hong Kong, China, to understand differences and implications for evaluation outcomes and to highlight possible CBA improvements. All countries studied used multicriteria analysis and examined capital, operating, and maintenance costs, and some included asset residual value. Monetized benefits varied, although travel time savings (TTS) were common primary benefits. Several countries also captured TTS for trucks, pedestrians, and cyclists. Accident cost savings (ACC) were common, although unit accident costs varied. For secondary benefits, all countries and regions except Hong Kong and Singapore included environmental externalities. Air pollution and noise impact were common. The United States was unique in including option value, and Germany and the Netherlands were unique in adopting agglomeration benefits. The social discount rate (SDR), assessment period, and decision criteria varied. Most SDRs used the marginal rate of return on private-sector investment (yielding an SDR of 6% to 10%). Net present value was the common decision criterion, and 20 to 30 years was a common evaluation period. Standardized parameter valuations suggested commuting value of time as $5 in Australian dollars per hour (A$5/h) to A$15/h, with an average of A$10/h for public transport users and A$10.50/h for car users. Accident cost varied; fatal accidents cost A$0.1 to A$4.25 million and serious accidents A$60,000 to A$490,000. A case study illustrates implications of adopting approaches with varied outcomes. Only Australia, the United States, the United Kingdom, and the Netherlands had positive benefit–cost ratios (1.00 to 2.61). TTS and congestion relief were major benefits (50% to 60% and 40% to 50%, respectively, of all benefits). ACC, environmental externalities, and option value benefits were not significant. Agglomeration benefits substantially increased project benefits.