The Colorado River does not reach the sea anymore. For most of the year, it runs dry somewhere in the Mexican desert well before it reaches the Gulf of California, absorbed entirely by the farms, cities, and industrial users who draw from it. This has been true, more or less, since the 1960s. The delta that once supported one of North America's most productive wetland ecosystems is now mostly dust. Tribes that fished and farmed there for centuries are gone. The people who divided the river did so in 1922, in a year that turned out to be among the wettest in the past five centuries. The Colorado River Compact allocated more water than the river actually produces in an average year, let alone a dry one. Everyone involved has known this for decades. The states that depend on the river have been arguing about how to address it ever since, an argument that has grown more urgent as the climate that feeds the river's snowpack has warmed and the cities and agriculture that rely on it have grown. The crisis that federal officials have been warning about for years is no longer theoretical. Lake Mead and Lake Powell, the two enormous reservoirs that serve as the river's storage system, have spent extended periods in recent years at historically low levels, triggering mandatory cutbacks to users in Nevada, Arizona, and Mexico. The agriculture-heavy Imperial Valley in California, which holds some of the most senior water rights on the river, has thus far avoided the worst cuts but faces a longer-term reckoning over whether the crops grown there, many of them water-intensive field crops like alfalfa and cotton, make sense in a desert fed by a diminishing resource. The politics of Western water are as old as the West itself, and they have always been contentious. Prior appropriation doctrine, the legal framework that governs water rights across most of the arid West, is built on a simple principle: first in time, first in right. Senior rights holders get their water before junior ones, regardless of who has the better use for it. This made a certain kind of sense in a frontier economy. It makes considerably less sense when a farmer with century-old rights is growing alfalfa in the Sonoran Desert while cities a hundred miles away face supply constraints. Reforming water law in the West is enormously difficult. The rights are property rights, and changing them means taking something away from someone who paid for it or inherited it. The political coalitions that have protected agricultural water use are durable, built over generations of relationship between farming communities, rural legislators, and the irrigation districts that serve as the institutional backbone of Western agriculture. But reform is happening, slowly, in pieces. The federal government has spent hundreds of millions in recent years paying farmers to fallow fields and leave water in the river, essentially buying conservation through compensation rather than forcing it through regulation. Some agricultural water districts have begun to trade water more actively, moving it from lower-value to higher-value uses through market mechanisms that the prior appropriation system was never designed to accommodate. Cities have done more than is generally recognized. Las Vegas, which depends almost entirely on Lake Mead, has reduced its per capita water consumption dramatically over the past two decades through aggressive conservation programs, turf removal incentives, and restrictions on outdoor water use. Tucson has invested heavily in water recycling and aquifer recharge. Phoenix is exploring desalination partnerships with Mexico. None of this is sufficient to close the gap between the water that is available and the water that current commitments assume. The math does not work. At some point, the states that share the Colorado must agree to use less of it, which means some combination of reducing agricultural production, limiting urban growth, paying for massive conservation, or accepting that some existing rights will not be honored. The federal government's authority to compel a resolution is real but politically costly to exercise. The Biden administration threatened to impose cuts unilaterally if the states could not agree, a threat that produced a short-term deal but did not resolve the underlying structural disagreement. The states are back at the table, negotiating a longer-term framework that will govern the river's management through the 2040s. The negotiations are difficult. The stakes for the forty million Americans who depend on the river are very high. Western water is, in this sense, a preview of a category of problems that the United States has not yet fully learned to manage: resource constraints that are clear in the aggregate but that run directly into established property rights, political constituencies, and institutional inertia that make rational collective action extremely hard to achieve. The river is smaller than the commitments made to it. How that gap gets closed will say a great deal about whether American governance is capable of addressing long-term problems before they become acute emergencies.