The authors review a number of recentinnovative rail concessions. Each country has approached itsproblems differently, and each provides different insightsinto what can be achieved with concessions. But all thecases show that restructuring and substantial governmentinvestment in the design of a concession pay off.Concessionaires can do exactly what is expected-increasetraffic, improve service, and enhance labor and assetefficiency if they are allowed to do so. A growing number ofcompanies and consortiums are interested in investing inrailway concessions if the concessions are offered onreasonable terms. And because both "positive"(where the concessionaire pays the government an agreed sumfor the concession rights) and "negative" (wherethe government pays the concessionaire to operate andmaintain the property) concessions are possible, loss-makingbut socially necessary services can also be concessioned.Perhaps the most important innovation in railwayorganization over the next few decades will be in theEuropean Union. Regulatory changes have ignited a cleartrend in the EU toward institutional separation ofinfrastructure from operations because infrastructure isseen as a state responsibility while operations (except forsocial services) are seen as commercial. One eventual resultof institutional separation will be franchising or evenprivatization of most freight services and possiblyintercity passenger services.