Three and a half years later, the promise of democracy in Iraq seems a cruel joke, and the administration in Washington that launched the invasion seems utterly bereft of clues as to how to sort out the mess. "The fall of the statue," says a former White House National Security Council staff member, "is where the story diverges from the White House's expectations."

President Bush's acceptance of Defense Secretary Donald Rumsfeld's resignation was intended to signal a change in course on Iraq. But many national security experts say that Condoleezza Rice, Bush's national security adviser at the time of the invasion, and the National Security Council share much of the blame for the problems in Iraq. "She did not perceive, and the National Security Council did not assess, what is in the United States' interests and what is in the interests of our enemies," says retired Lt. Gen. William Odom, an outspoken critic of the war who served as military assistant to Zbigniew Brzezinski, President Jimmy Carter's national security adviser. "Once you make that basic mistake, there isn't any way to make the war come out good. It's all over."

With her ardent support for the invasion, Odom and others say, Rice was unable to play the traditional role of national security adviser-impartial broker in the rough-and-tumble of interagency government. That was further complicated by the fact that Rumsfeld all but ignored the input of other agencies, then never took responsibility for postwar reconstruction, leaving Rice to try to manage the Iraq rebuilding effort from the White House. "The people who took responsibility for the rebuilding of Iraq-that is, the office of the secretary of defense-expressed no interest in solving problems as problems appeared," says Franklin Miller, one of Rice's principal aides handling the reconstruction effort. "They never brought any issues to the table, and they never took any taskings away."

The problems were compounded, former NSC staff members say, by conflicting information provided by virtually everyone involved in the Iraq effort, including military generals who often gave NSC staffers information at odds with what soldiers on the ground and lower-level military and civilian officials were reporting. "The internal whitewashing was as great as the external whitewashing," says a former senior NSC staff member. "We wasted two years, spinning the situation, using smoke and mirrors to say, 'Hey, it's not as it's being reported. Everything is OK.'"

"Knocking heads." This is the story of how Rice, a veteran White House insider with close personal ties to the president, failed to pull together the administration's fractious postwar efforts and why her successor, Stephen Hadley, has yet to navigate or even identify a way out of Iraq. "Hadley's NSC," says a former senior NSC staff member, "has become a rubber stamp for the Defense Department." Bush has tasked Hadley, Rice, now secretary of state, and other advisers to come up with a solution for Iraq, even as the congressionally directed Iraq Study Group, cochaired by James Baker, the former secretary of state, is racing to finish up its own report. Completing the trifecta is a report to be issued next month by the Pentagon on U.S. military strategy in Iraq.

An extensive review of the National Security Council's role in the Iraq war and its aftermath-based on interviews with a dozen former NSC staff members, senior officials from the State Department and the Pentagon, and outside security experts-reveals new details of the White House's failure to effectively manage the interagency effort on Iraq. The review by U.S. News is based on a detailed recounting of the NSC's interagency deliberations, including the daily and weekly secure video teleconferences conducted from the White House's Situation Room. Rice and Hadley declined to be interviewed for this account. Rumsfeld failed to respond to an interview request.

Several important new books, including Fiasco, by the Washington Post's Pentagon correspondent, Thomas Ricks, and State of Denial, by Bob Woodward have chronicled the many missteps in the Bush administration's prewar planning and post-invasion conduct. But historians will be discussing for years how a venture vested with such singular importance by its advocates could have been conceived and executed with such myopia and ineptitude. To date, the administration has failed to meet a single one of the reconstruction goals it set for itself back in 2003-goals for rebuilding infrastructure, defraying reconstruction costs by increasing oil production, and instituting a constitutional democracy with functioning courts and the rule of law.

To understand the genesis of the problems in Iraq, it is helpful to know a bit of the history of the position to which Rice was appointed in January 2001. Congress created the National Security Council in 1947 to serve as the president's principal forum for resolution of military and foreign-policy matters. The council has four permanent members: the president, the vice president, the secretary of defense, and the secretary of state. A variety of other cabinet secretaries and military and other presidential aides attend NSC meetings, as well. The job description of the national security adviser has been shaped over time by presidential fiat rather than by statute, and while it has been conceived somewhat differently from administration to administration, it has generally been viewed as the president's primary surrogate for bringing dueling cabinet secretaries into line, resolving differences, and framing issues for decision. The term "honest broker" has typically been affixed to the job description. So has the term "knocking heads together." "Rice seemed to take the position from the very beginning that she wasn't going to knock heads," says Robert Perito, a retired career Foreign Service officer, who served on President Ronald Reagan's NSC staff and is at the United States Institute of Peace. "She was just going to facilitate" for the president.

"Stovepiped." The result, virtually every close observer of the Bush White House agrees, was a power vacuum that was filled by Rumsfeld and Vice President Dick Cheney-particularly after the 9/11 terrorist attacks-with Secretary of State Colin Powell edged to a position of just marginal influence. "For a long time, DOD got whatever they wanted," says a former NSC staff member, using the bureaucratic shorthand for the Department of Defense, "because they were winning [in Afghanistan] and because Cheney backed them."

On Iraq, the Cheney-Rumsfeld axis was stronger than ever. The two old friends were convinced of the urgent need to remove Saddam; views to the contrary were unwelcome. Even in the most intimate meetings with the president in the Oval Office, where one was assured of confidentiality, says a former NSC staff member, there was little candor and few conflicting views, including from Rice. "The discussions of how things were going were so wildly optimistic and so out of line with the intelligence," this staff member says, "it was almost a cheerleading role."

In December 2002, President Bush made a crucial decision, to give the Defense Department, rather than State, responsibility for postwar reconstruction in Iraq. Rumsfeld pushed hard for the responsibility, says Douglas Feith, one of the defense chief's top aides in the war-policymaking effort, because he didn't want to see a recurrence of the problems that arose in the Balkans, where the responsibility for civil reconstruction and security operations was under separate chains of command. "The people responsible for civil reconstruction [in the Balkans] said, 'We can't take the troops out or everything will collapse,'" says Feith. "So the troops became hostage to the lack of progress on infrastructure reconstruction."

In January 2003, President Bush signed a national security presidential directive, NSPD24. It was drafted by Feith and by Hadley, Rice's deputy. The presidential order gave the authority for management of postwar security, stability, and reconstruction in Iraq to the Pentagon and created the Office of Reconstruction and Humanitarian Assistance. At Rumsfeld's request, Feith called Rumsfeld's friend, retired Army Gen. Jay Garner, and asked him to head up ORHA and begin the postwar planning for Iraq. But Garner says he was told he probably wouldn't ever actually go to Baghdad because the president wanted to name an envoy who had national recognition. "I went in as a temp," Garner told U.S. News. "That was always known in the administration." Garner had just five weeks to coordinate the postwar plans of all the different U.S. government agencies, which, he says, had all been "stovepiped"-effectively, shoved into different drawers, one having no relation to the next.

In February, the U.S. Army War College released a study by Conrad Crane and W. Andrew Terrill spelling out the urgent need for serious interagency planning, a multiyear commitment to Iraq, and a real nation-building plan. "I understand there were competing truths out there," Crane says. "I believe the CIA had much more rosy scenarios than we did. But people within the Army took the study seriously. Many planners took it seriously."

The assault on Iraq began in March 2003, with the 3rd Infantry Division the first American troops to reach Baghdad. Not long after the statue of Saddam fell, U.S. troops found themselves battling with insurgents almost daily. "All of us had military people coming to us and saying, 'This isn't going the way they are saying it's going to go,' " says Kori Schake, a former staff member at the NSC who once worked at the Pentagon and now teaches at West Point. "From the bottom up, everybody was saying, 'What happens once we have Iraq?'" says Schake. "How does a military plan that emphasizes speed and a city-skipping strategy translate into control of the country?"

At meetings in December 2002 and January 2003, NSC sources say, Bush asked Gen. Tommy Franks-the commander of the Central Command, which drafted the Iraq war plan-how the United States would establish security after the war there. The military would leave "lord mayors" in every city, Franks told Bush-although it was never clear precisely what Franks meant. The lord mayors, in any case, never materialized. Franks did not respond to an interview request.

"Lame duck." On April 18, 2003, tens of thousands of Iraqis protested the U.S. occupation of Baghdad and marched in favor of an Islamic state. The looting and burning were so serious that Franks refused to allow Garner into Baghdad until April 21. By then, the city had been stripped bare. On April 24, Garner's first full day in Baghdad, the retired general was sitting alone at a desk in a dusty room in one of Saddam's looted palaces when the phone rang. It was Rumsfeld. He told Garner brusquely that Paul Bremer had just been named U.S. envoy to Iraq. "What that means," Garner recalls thinking, "was that my first full day in Baghdad, with my entire team, I was the lame duck beginning that day."

For the next two months, Garner's team of 350 people had to begin reactivating Iraq's government ministries, print and issue new currency, fix power grids, restore water supplies, draw up a new legal code, repair bridges and roads, and get the schools back in business. The State Department had drawn up elaborate plans, but Rumsfeld and Feith ignored them. Garner's challenge, unsurprisingly, proved to be overwhelming. "We didn't have specific plans tested for years and red-teamed the way the military plans are," Feith concedes. "We hadn't integrated the plans with the combatant commands. We hadn't done deployment exercises." Feith argues that the U.S. government's postwar efforts have always been ad hoc. But what happens after the fighting's over? "Sometimes that goes well," Feith says, "and sometimes that doesn't go well."

On April 28, Fallujah, a key Sunni city and one of the most peaceful areas of Iraq, went up in flames after U.S. troops fired on an unarmed crowd marching in violation of curfew to protest the U.S. presence; 16 Iraqis died. On May 7, a week after President Bush's now famous "mission accomplished" visit to the USS Abraham Lincoln, senior military officers gave Rumsfeld an "upbeat" assessment of the situation in Iraq and began making plans for drawing down U.S. troop levels there by September 2003, according to a report from the Center for Naval Warfare Studies. By June, that same report concluded, central Iraq was "in the midst of a low-level, decentralized insurgency."

On May 12, Paul Bremer took over as the interim Iraq administrator, and Jay Garner's ORHA team was dissolved into Bremer's new Coalition Provisional Authority. Bremer's decision, in mid-May, to outlaw the top tiers of Saddam's Baath party and dismiss some 30,000 Baathists from various ministries has been widely criticized as among the worst decisions of the postwar phase. But de-Baathification, Bremer says, was the correct policy. "The mistake I made," he explained in an interview, "was putting the implementation of that [policy] into the hands of the political body, the governing council, rather than a judicial body, which would have been more objective. Once it got to the politicians, they implemented the policy in a far broader scope than we intended, and that was a mistake."

On May 23, Bremer disbanded Iraq's 500,000-member military and intelligence services, leaving thousands of angry, jobless men with guns on the streets. "If we had reinstated Saddam's Army, which had been the key instrument of his repression," Bremer told U.S. News, "it would have led to an immediate civil war." Key military officials, including Gen. Peter Pace, then vice chairman of the Joint Chiefs of Staff, later said that the Joint Chiefs were not told in advance of Bremer's decision. Neither, NSC officials say, was Rice.

With the violence in Iraq escalating and the number of American casualties climbing, President Bush asked Rice to become more involved in the postwar effort. The White House insider who had left the power vacuum 15 months earlier was now being asked to fill it-immediately. "As the magnitude of misjudgments becomes visible in Iraq," says a former NSC staff member, "Condi becomes more responsible for the management of the process."

On June 18, Rice met with the NSC staff and expressed her fears about the consequences of Bremer's decision to disband the Iraqi Army. Were there enough U.S. troops in Iraq to deal with a further upsurge in violence? she asked. "It looks like we're not in a stability phase. It looks like we're still in a combat operations phase," Rice said, according to a former staff member. "It looks like we're still trying to impose our will."

Rice told Miller, her senior director for defense policy and arms control, that she wanted him to restart the Executive Steering Group, an interagency group he had cochaired back in the summer of 2002, when the early war planning had begun. Rice told Miller, "Find out what we have to do," an NSC staff member recalls, "and find out how much time it will take and how much it will cost."

Each morning at 6:30, Miller rushed into his office in the Old Executive Office Building next to the White House, peeled off his coat, rolled up his sleeves, and grabbed the phone. How much power had the grids in Iraq generated overnight? What about the water? What was yesterday's death toll? In an interview with U.S. News, Miller recalled that the NSC staff was deeply engaged in postwar efforts, but he downplayed his role and that of his staff. Some in the Pentagon complained of White House micromanagement, but Rice made it clear to her staff she wanted action. "Dr. Rice reconvened the group because things weren't working," Miller says, "so we got very operational. That's because nobody else was doing it. We were fighting fires every day."

On June 20, Army Col. Anthony Harriman, the Pentagon's assistant deputy director for Joint Operations, told NSC staffers that the Joint Chiefs believed the level of combat forces in Iraq was sufficient and that the plan was to take the military police and artillery regiments and have them operate as "straight-leg infantry," NSC sources say. "What they recognized was that things weren't quite what they should be," says one staff member. "But they thought, 'We can take people trained to do different things and make them act as infantrymen. That'll take care of the problem.'"

It didn't. While some members of the Pentagon's Joint Staff were talking about drawing down troop levels, other information coming to Miller and the NSC staff indicated that things in Iraq were spiraling out of control. "In many instances, we'd be reporting what the generals were saying, [and] we thought we were giving Condi a measure of truth," says a senior NSC staff member. "But what was happening was the generals were also putting a big spin on it."

"Sanity checks." If anyone on the NSC staff could get hard information, it was Miller. A career civil servant, he had spent 31 years in the government, including two years at the State Department and 22 years working for seven secretaries of defense. Miller had been awarded the Pentagon's highest civilian medal five times as well as the department's distinguished public service medal.

Miller's Iraq team had superb Pentagon connections. NSC staff members made furtive forays onto the military's secure intranet, the SIPRNET, for data about troop casualties and insurgent attacks and started compiling their own spreadsheets. They also tapped sources at Franks's Central Command or got sympathetic sources at Bremer's CPA or former colleagues at the Pentagon to let them peek at computer screens to see what new information was coming in from Iraq, and they called friendly embassies for "sanity checks." One reliable source was Deputy Secretary of State Richard Armitage, who shared the NSC staff's concerns. A veteran defense official who had served three combat tours in Vietnam, Armitage used his tentacles at all levels of the military and the media to get real data and share them with the NSC staff.

Members of Cheney's office also proved to be allies, if unlikely ones. Contrary to much of the conventional wisdom, the vice president had only a small national security staff working on Iraq, and-unlike in the run-up to the invasion-he played very little role in the postwar efforts, NSC sources say. The two Cheney officials most steeped in Iraq issues were career Foreign Service officers, both willing to help. The NSC staff then started asking DOD questions. "The military staffs began to respond," says a former staff member. "But they knew we were passing this information to Rice. So they would report the numbers but downplay the significance of the trends."

Each week, there were scores of White House meetings on Iraq in which the NSC staff participated. Gradually, Rice shifted from the counselor role she had envisioned herself playing to more of a coordinator. "This was the president's priority," says a former NSC staff member, "and the NSC staff was being the sheepdog or border collie, yipping at people to make sure things were getting done." Executing was Miller's job. "I can't tell you the number of times," says a former staff member, "when Dr. Rice would turn to Frank and say, 'Frank, I want you to fix it.'"

But there was only so much even a seasoned veteran like Miller could do. Rumsfeld was a consummate infighter with a strong aversion to briefings that didn't have "some large self-evident truth," says Feith, a former deputy. "Every briefing has to say, 'Here's the cosmos; here's where we are in the cosmos.' You need a map of the cosmos, with a sign saying, 'You are here.'" Once that was established, Feith says, Rumsfeld wanted to spend much of the debate on assumptions that may have been made about the cosmos. "His entire approach to life," says Feith, "is that whatever assumptions you have must be re-examined constantly in the hard light of facts."

What that meant for the NSC interagency process, staff members say, was that very little got done, because every meeting went back to "first principles"; decisions about second-tier issues-things like reconstruction-never got made.

"We got it." In Baghdad, the new head of the Central Command, Gen. John Abizaid, believed it was imperative to get the kids with guns off the street. His solution: the Iraqi Civil Defense Corps. "They way it was portrayed to us," says a former NSC staff member, "was, 'It's a great thing.'" But the program was neither funded nor adequately planned out, Rice discovered. U.S. troops were told to go to street corners and recruit young Iraqis, who were happy to join the Civil Defense Corps-until they had to actually fight the insurgents. "They fell apart," says the former NSC staff member. "They ran away; it was untenable." Bush pressed for answers. "I saw the president ask Rumsfeld, Powell, and Abizaid at the biweekly NSC meetings on Iraq, 'Is there anything I can do?'" the former staff member says. "Nobody spoke up. It was 'No, sir, we got it.'"

On October 2, David Kay, the U.N. inspector charged with hunting down Saddam's weapons of mass destruction, delivered a devastating report. After an exhaustive four-month search, he said, no such weapons had been found. A day later, a New York Times/CBS poll showed that support for Bush's Iraq policy was at rock bottom.

On October 6, Rice established the intra-NSC Iraq Stabilization Group. Its mission: Try to stop the train wreck. Each week, Miller and others on the staff briefed Rice, Hadley, and members of the ISG about the military, Iraqi security forces, stability operations, and reconstruction efforts. NSC staff member Schake briefed Rice on coalition management of troop-contributing countries. Gary Edson, the senior director for international economics, briefed the group about Iraqi economic issues. Frances Fragos Townsend, Bush's homeland security adviser, addressed counterterrorism efforts. Rice had also reached out to an old boss, Robert Blackwill, who had recently served as ambassador to India, to come work at the NSC, and he quickly became focused on Iraqi politics and governance issues.

The various ISG subgroups began attacking a whole range of crises. Miller's group took security, stability, and urgent reconstruction tasks, dubbed SWET-sewage, water, electricity, trash collection. "If we had provided palpable gains in those areas," says Miller, "the Iraqis would have said, 'The future is getting better.'" This later became known as the "man on the moon" issue-Iraqis were saying, if the U.S. government could put a man on the moon, why couldn't it provide them the most basic amenities?

Rice's staff knew the Iraqis were right.

The first task at hand was ensuring that Iraqis had enough rice and beans, for the holy season of Ramadan. "The notion that the White House staff had to get involved with beans and rice is almost laughable," says Miller, "except for the consequences if it hadn't happened. And it wasn't going to be fixed without top-down guidance."

The next issue was power. After the United States took over Iraq, power generation averaged 3,900 megawatts. Bremer had indicated to Rice that by October 2003, power generation would be up to 6,000 megawatts a day. The goal was never met. Miller's objective was to force electricity generation higher. When the Army Corps of Engineers ran out of money to upgrade nearly 150 electrical generation facilities, the NSC staff and the U.S. Office of Management and Budget found some money and got the plants upgraded. When workers refused to show up to install the generators, NSC staffers demanded that Bremer's CPA find a way to house them on-site.

Then there was infrastructure-Miller arranged to have the OMB transfer $250 million in reconstruction funds to support efforts by an Army lieutenant general named Peter Chiarelli in Baghdad, who was paying 18,000 residents in Sadr City to help rebuild their neighborhood. "It shouldn't have been the NSC staff," says Miller. "It should have been DOD and the [U.S.] Embassy."

In the summer and fall of 2003, the other big challenge for the NSC-and Blackwill, in particular-was how to help the Iraqis move to a working democracy and help them pull off an election. Rice and Blackwill were old colleagues-he was her boss and mentor while both served on George H. W. Bush's national security staff, specializing in Russian and eastern European issues. Blackwill was viewed as brilliant but prickly. He also had the reputation of having difficulty dealing with women. But Rice trusted Blackwill and knew he had the right stuff for the job. He was also an old friend of Bremer's.

Accounts differ over precisely what happened next, but suffice it to say there was intense disagreement among those assigned to the NSC staff over the best way to try to pull Iraq back from the precipice. Rice told Blackwill she wanted him to pursue a strategy to bring disaffected Sunnis back into the fold, in the hopes of quelling at least some of the violence. Around that same time in Baghdad, a young and upcoming aide to Bremer named Meghan O'Sullivan was forging deep ties with leaders of the Shiite political party known as the Supreme Council for Islamic Revolution in Iraq, or SCIRI. In June 2004, after Bremer handed over sovereignty to the Iraqis and transitioned the CPA over to the new U.S. Embassy in Baghdad, Blackwill offered O'Sullivan a job back at the NSC. But Blackwill's tenure there was short-lived. He was replaced by O'Sullivan. In Baghdad, the young national security aide had worked to establish closer ties to the Shiite political leadership, including Adel Abdul-Mahdi, a moderate who ranked as the senior SCIRI member in the Iraqi government.

The association resulted in some tense moments. During one meeting with Rice, according to numerous NSC sources, O'Sullivan advocated strongly for the Bush administration to support Shiite candidates in the national assembly elections, set for January 2005. Visibly disturbed, Rice delivered a long lecture on constitutional democracy, explaining why it would be inappropriate for Washington to intervene on behalf of any individual Iraqi political party. In the end, says one staff member, when O'Sullivan realized that Rice wasn't going to budge, she reversed course. But the incident, several NSC staff members say, shows how contentious things were even within the small, relatively cohesive White House staff. In the view of some, O'Sullivan was primarily interested in encouraging the Shiites to form a moderate coalition with the Sunnis. Others complained that she repeatedly advanced the view of SCIRI members. Still others argued that Washington should make every effort to avoid getting embroiled in the internecine Iraqi political scene. The contentiousness among the NSC staff roughly reflected some of the same differences within the upper levels of the administration. Cheney's office, for instance, was for near-absolute noninvolvement. Rice, on the other hand, thought it made sense to quietly push for a moderate coalition. In the end, such gaps proved largely unbridgeable.

"Please, fix it." In the spring and summer of 2004, Rice and the NSC had oil on their minds. Rice was concerned that without fuel, the electricity generators would not function, and that without generators, there would be no air conditioning in the 110-degree summer heat. She also feared that insurgents were trying to isolate and bring down the Baghdad government by attacking the oil pipelines. In the end, the government failed to make inroads to boost oil production and refining, in large part because of the disarray of the CPA, the ineptitude of the Iraqi ministries, lack of foreign investment, and the sophistication and relentlessness of the insurgents' attacks. Rice was more frustrated than ever. "You could just see that she was tired," says a former NSC staff member. "And when we would come to her, she would just wave her hand and say, 'Please, fix it.'"

One of the most useful tools available to U.S. forces in Iraq was the Commanders' Emergency Response Program fund. The CERP money was given directly to commanders and tactical forces so they could quickly pay Iraqis for services and to rebuild infrastructure. In March 2004, Lt. Gen. Ricardo Sanchez, commander of the coalition ground forces, asked Bremer to release more CERP funds. Bremer refused. Sanchez went up the chain to Rumsfeld's deputy, Paul Wolfowitz, who appealed to Bremer. Again he refused. Soon after, the NSC's Miller was in Iraq and learned about the problem; he told Rice, who made Bremer release the funds.

The NSC staff began calling Bremer the "Last Viceroy." "He was the proconsul. We all existed at a level way below his," says an NSC staff member. "And he instilled in his people the fear of God that said, 'Don't do anything that Washington tells you to do, and you don't do anything unless I tell you to do it.'" Bremer told U.S. News that he reported to the president through Rumsfeld, and that starting in October 2003, he spoke with Rice virtually every day. "I did object to third-string staff people in Washington second-guessing decisions we were making," Bremer told U.S. News, adding: "My view was if anybody back here had a really good idea about how to do things better, I had a flak jacket, a helmet, and a cot available for them."

Bremer's authority, along with his and Rumsfeld's micromanaging, made it impossible for Rice and her NSC staff to win the David-versus-Goliath struggle against the Pentagon and CPA bureaucracy. "I think they often wanted to do the right thing," says a former Pentagon official. "They didn't always put their muscle into the right thing. When they did, it would be helpful. But often, it wasn't enough."

Starting in the summer of 2004, one of the most intractable issues for the NSC staff was getting realistic assessments of U.S. efforts to train the Iraqi police and military, without which the United States could not even contemplate its own exit strategy. The job of troop training was given to one of the military's brightest stars, Lt. Gen. David Petraeus, who had won plaudits for leading the 101st Airborne Division during the Iraq invasion and the subsequent occupation of Mosul, into early 2004. At the time, Petraeus had been credited with fixing up the schools in Mosul, getting the phones to work, the clean water to flow, the lights turned back on, and generally making life easier and safer for Iraqis. But there was concern in Washington that these seemed to be very temporary fixes, and that no sooner had the 101st Airborne taken credit for cleaning up a section of Mosul and left than the violence erupted again. "He was saying how effective he was being," says a former NSC staff member, "but throughout 2004, the CPA and, later, the embassy were reporting that Mosul was a mess."

Petraeus told U.S. News that he offered realistic assessments up the chain of command. "Mosul did quite well there, for the bulk of our time," Petraeus says, "even though we had a tough time at various points, especially in November of 2003, during Ramadan." The 101st Airborne was redeployed home by the end of January. Petraeus left Iraq in mid-February.

The 101st Airborne was replaced with an assortment of other units, which was "half of what we had, and by November, it was at a third," says one Army officer. "And the new unit did not have the robust intelligence analysis structure that we did earlier." By the summer of 2004, Mosul had become a fertile ground for insurgents again.

"Metrics with rigor." In June 2004, Petraeus was tasked with training Iraqi security forces, a crucial responsibility because it was tied directly to the U.S. exit strategy. Once again, the initial news was good-Rice's aide, Blackwill, returned from a trip to Iraq that September and reported that Petraeus was pumping out lots of newly trained Iraqi security forces; Petraeus himself said he was making "huge progress." But by December Blackwill's assessment had become considerably bleaker. In January 2005, the State Department issued a report giving Iraqi security forces a failing grade; Pentagon officials now acknowledge that they have no idea how many of the more than 300,000 U.S.-"trained" troops are still in the Iraqi security forces. Petraeus says he was always candid in his assessments of Iraqi troop readiness. "We tried very hard to develop metrics with rigor," he says, "and to continually improve them."

In January 2005, Rice was confirmed as Bush's new secretary of state. Four days later, an estimated 8 million people in Iraq went to the polls to vote in elections for a transitional national assembly. The United Islamic Alliance-a coalition of mainly Shiite groups-won a majority of assembly seats, followed by the Kurdish party.

Back in Washington, President Bush asked Rice's former deputy, Hadley, to become his national security adviser. Once he took charge, Hadley's message was crystal clear: "The NSC doesn't do operations." He immediately pulled the NSC staff back from its "border collie" role in Iraq and shut down its remaining efforts on reconstruction, infrastructure, oil, and water. By then, Miller and many of his colleagues working the Iraq issues on the NSC Iraq staff had left for the private sector. Last November, the NSC released a major report, "National Strategy for Victory in Iraq." This past summer it was roundly criticized by the nonpartisan Government Accountability Office, the watchdog arm of Congress. "We said it was lacking in important respects," says the GAO's Joseph Christoff, who did the evaluation. "It doesn't say how much it will cost; it doesn't say who will take the lead."

The answers to those questions remain elusive to this day. Perhaps they will be provided by one of the three Iraq studies now underway. Or perhaps not. "What's frustrating," says one former NSC staff member, "is that we're dealing with one small country, and we're tying ourselves in knots."

To contact the author: cragavan@usnews.com

With Danielle Burton and Stephanie A. Salmon