Call it the "Anaconda Strategy." That's the moniker Bush administration officials gave their fallback plan for dealing with the radical cleric Moqtada Sadr after their crackdown in early April triggered a revolt by Sadr's followers. It may be good shorthand for the administration's cut-and-paste approach as it heads toward next week's return of sovereignty to Iraq.

America's make-do policy in the Iraqi endgame was outlined by a senior administration official. From his account it's clear that Bush and his advisers have been improvising for the past few months, struggling to craft an exit strategy. The grand designs that launched the war are now long gone, replaced by a process of trial and error.

The first challenge this spring was Fallujah. When Sunni Muslim insurgents there brutally murdered four U.S. civilian contractors in early April, the White House ordered U.S. Marines "to go get the killers," the official said. But the Marine offensive produced heavy civilian casualties in Fallujah, rage among Iraqis, warnings from Jordan and other neighboring states, and a resignation threat by several members of the Iraqi interim Governing Council.

Three days into the Fallujah offensive, "the president made a decision we should pause and see how we could do without the Marines going in," the senior official said. The Marine commander there patched together a "Fallujah Brigade," headed by a retired Sunni general in the Iraqi Republican Guard, to restore order. The fighting eased, but Fallujah remains a stronghold for insurgents and a threat to the new Iraqi government.

A second challenge was the uprising by Sadr and his Shiite Muslim followers. The administration provoked the confrontation in early April by closing Sadr's newspaper and arresting his key lieutenant. Sadr's followers responded by seizing control of parts of southern Iraq, including the holy cities of Karbala and Najaf. The administration considered an all-out attack but then pulled back. "The effect on Shiites in Iraq, and beyond, of the U.S. military taking down their two holiest sites was daunting, to say the least," the official explained.

The new "Anaconda Strategy" was a compromise that specified attacking Sadr's militia everywhere but in Karbala and Najaf. "The president's view was, 'Keep the pressure on, but don't go into those cities,' " the senior official said. Two months later the administration regards this strategy as a success. Hundreds of Sadr's followers have been killed, and Sadr himself said last week that he planned to enter politics.

A bizarre sidelight was Iran's offer in the midst of the Sadr uprising to serve as a mediator. An Iranian would-be negotiator went to Najaf, hoping to meet with Grand Ayatollah Ali Sistani. But after Sistani refused to see him, the Iranian effort collapsed.

U.S. officials believe they finally have reliable communications with Sistani after a year of botched contacts through intermediaries. The communications now are almost entirely in writing, which suits Sistani's clerical penchant for textual analysis.

The final challenge in the run-up to Iraqi sovereignty has been the selection of an interim government. This process was jointly managed by U.N. special representative Lakhdar Brahimi, Bush National Security Council official Robert Blackwill and U.S. occupation chief L. Paul Bremer.

In choosing an interim president, the three "B's" in Baghdad forwarded to Washington the names of two Sunnis on the Governing Council: Adnan Pachachi and Ghazi Yawar. The White House responded that either was acceptable. Hoping to boost Yawar's popularity with Iraqis by distancing him from Washington, his supporters leaked to American journalists a false report that Bush had decided on Pachachi. In fact, it was Brahimi who wanted Pachachi. When the 80-year-old Pachachi turned the job down, it went to Yawar by default.

Ayad Allawi became prime minister somewhat by default, too, the official said. Washington was enthusiastic about Brahimi's first choice, a Shiite nuclear scientist named Hussain Shahristani, and found nothing in his past that would disqualify him. But he was little known in Iraq, had no support within the Governing Council and thus failed the "market test," the U.S. official said.

Allawi seemed the strongest alternative Shiite candidate, the official said, because he had support among Sunnis and Kurds and the endorsement of Sistani. Three senior members of the Governing Council backed Allawi, and when the council as a whole was informed that he was the choice, some members falsely claimed they had elected him.

The Iraq transition has been a haphazard, stop-and-go affair. Because of poor planning and missed opportunities, Washington had to throw away its playbook. The chief U.S. accomplishment may be the simple fact that in just over a week, Allawi will be in charge, not Saddam Hussein.