FOUR YEARS AGO, HIS WORDS WOULD have represented an almost unquestioned consensus view. In late January, the State Department's counterterrorism coordinator, Dell Dailey, described al Qaeda's top leadership as isolated, saying that they have "much, much less central authority and much, much less capability to reach out."

He is not alone in this assessment. In July 2007, Stratfor's Peter Zeihan argued that while a few thousand people may claim to be al Qaeda members, "the real al Qaeda does not exercise any control over them. . . . The United States is now waging a war against jihadism as a phenomenon, rather than against any specific transnational jihadist movement." The most prominent proponent of this view has been Jason Burke, a reporter for London's Observer and the author of Al-Qaeda: The True Story of Radical Islam. By the time that book hit newsstands in 2003, Burke was already arguing that the "nearest thing to 'Al-Qaeda,' as popularly understood," only existed for a five-year period, and the battle of Tora Bora in December 2001 showcased "the final scenes of its destruction." Now, Burke contends, we are "in a 'post-bin Laden' phase of Islamic militancy."

Unfortunately, all these men are wrong--and we will fight the war on terror less effectively if we continue to harbor mistaken assumptions about the al Qaeda network. It is important not to overstate what the terror group's leadership needs to do to remain relevant. Even if the central leadership's role is limited to connecting terrorist nodes--pairing skill sets, financing, and operatives--it can transform terrorist groups from disunited regional problems into cohesive adversaries capable of threatening Western societies. Moreover, the safe havens that al Qaeda's leaders have gained in recent years magnify their lethal capabilities.

Al Qaeda itself has faced internal debates about its future. Abu Musab al-Suri, one of the most prolific jihadist ideologues, in recent years has argued for a decentralized combat model. In contrast, Abu Bakr Naji, another prominent ideologue, calls for a more centralized model.

Suri's 1600-page manifesto, The Call for Global Islamic Resistance, argues that the centralized, hierarchical model of jihadism cannot overcome the U.S.'s technologically advanced military, and that regional security cooperation--such as the alliance between Washington and Islamabad--makes a hierarchical structure dangerous. He suggests that decentralization immunizes terror cells from detection through the capture and interrogation of members of other cells. Suri's prescription for decentralization would mean replacing the old training camp model with one in which fighters are trained "in homes and mobile camps."

In contrast, Naji's The Management of Savagery argues that once the jihadists hold territory, they should erect a governing apparatus to enforce Islamic law and provide security, food, and medical care. A high command would ensure that efforts are not needlessly duplicated, and would prioritize actions against various groups or nations. Naji's argument has carried the day within al Qaeda's hierarchy. Though there are many reasons for this, perhaps the most significant factor has been external events. As al Qaeda gained new safe havens in Pakistan and beyond, Naji's model seemed most fitting.

External events aside, the preference of al Qaeda's leadership for Naji's approach over Suri's reflects a long-standing inclination for centralization. Osama bin Laden originally formed al Qaeda to keep the vanguard of jihad alive after the Soviet Union's defeat in Afghanistan. West Point's Combating Terrorism Center has translated a number of documents captured during the Afghan and Iraq campaigns that the Department of Defense has declassified from its Harmony Database. These documents depict a clear al Qaeda hierarchy dating back to bin Laden's residence in Sudan between 1992 to 1996.

One document, entitled "Interior Organization," delineates al Qaeda's hierarchical structure, from the commander and ruling council down to organizational committees. It explains that the commander must have been a member of al Qaeda for at least seven years, have a sufficient understanding of Islamic law and jihad, and "have operational experience from jihad." The document also enumerates five separate committees: military, political, administrative and financial, security, and surveillance. Other documents detail members' duties, salaries, and even vacation time. Bachelors qualify for a round-trip ticket home after a year, although they have the option of using it for hajj (religious pilgrimage) instead. An application to train in al Qaeda camps inquires about the applicant's education level, professional experience, medical history, and how much of the Qur'an he has memorized.

Although the U.S. invasion of Afghanistan devastated al Qaeda's safe haven, the group's core leadership survived. A few--such as Saif al-Adl, Saad bin Laden, and Sulaiman Abu Ghaith--fled to Iran, but most relocated to Pakistan. Soon after, al Qaeda's regional nodes took the lead in operations. There were three such regional attacks in October 2002 alone. On October 8, 2002, two Kuwaitis linked to al Qaeda opened fire on U.S. marines, killing one. On October 12, Indonesia's Jemaah Islamiyya killed 202 people in a nightclub bombing in Bali. On October 23, Chechen terrorists seized a Moscow theater packed with 850 people. The March 2003 capture of Khalid Sheikh Mohammed, al Qaeda's principal operations commander and chief architect of the 9/11 attacks, set back the reassertion of al Qaeda's central leadership. Over the next year, regional attacks continued. We had grown so accustomed to attacks led by regional nodes that when the March 11, 2004, commuter train attacks in Madrid and the July 7, 2005, London suicide bombings were executed, they were immediately described as having little connection to al Qaeda's senior leaders.

With the benefit of hindsight, it is clear that in this period analysts and media commentators underestimated the extent to which al Qaeda's central leadership remained able to organize terror attacks. Although a regional node implemented the Madrid plot, al Qaeda's senior leadership formulated the operation. The Center of Mujahideen Services, an internal al Qaeda "think tank," developed the political strategy behind the attack in the book Iraq al-Jihad, which concluded that "the Spanish government will not endure two or three attacks." The book thus argued that a coordinated terrorist assault could turn the Spanish public against the government, forcing it to withdraw troops from Iraq.

There were also operational connections between the Madrid cell and the broader al Qaeda network, but the connections were more dramatic for London's 7/7 attacks. British police reports were actually hesitant to link the 7/7 bombers to al Qaeda, describing the terror cell as autonomous and self-actuating. But as the official account of the 7/7 attacks hit the British press, terrorism analysts Dan Darling and Steve Schippert enumerated a number of problems with concluding so early that the broader al Qaeda network was largely irrelevant to the London plot. They noted connections between cell leader Mohammad Sidique Khan and Riduan Isamuddin, mastermind of the Bali bombings. Mohammed Junaid Babar, a Pakistani native living in Queens, New York, who pled guilty in federal court to smuggling military supplies to al Qaeda and assisting in a UK bombing plot, had identified Khan as someone he had met at an al Qaeda camp in Pakistan. Haroon Rashid Aswat, who helped set up an al Qaeda training camp in Oregon, had telephoned the London bombers hours before the attack. After the bombing, Khan and fellow bomber Shehzad Tanweer appeared in a video aired on the Arabic satellite channel al-Jazeera that included praise for the attacks from bin Laden and his deputy Ayman al-Zawahiri, as well as Khan's suicide message. It is unlikely that al Qaeda's senior leadership would have this footage were they unconnected to the attack. Underscoring this point, al Jazeera aired a new video from Zawahiri on the first anniversary of the bombings claiming that Khan and Tanweer had visited an al Qaeda camp in Pakistan "seeking martyrdom." Bob Ayers, a security expert at London's Chatham House think tank, commented, "It makes the police look pretty bad. It means the investigation was either wrong, or they identified links but were reluctant to reveal them."

The connections between al Qaeda's senior leadership and the attacks in Madrid and London demonstrate that the group's top command was not as isolated and irrelevant during this period as some suggested. Still, it would gain more strength over time.

After relocating from Afghanistan to Pakistan, al Qaeda's senior leadership set about revamping its operations. They tried twice, in December 2003, to kill Pakistani president Pervez Musharraf, leading him to send troops into Pakistan's tribal areas. Al Qaeda and allied tribes prevailed in the fight.

The 9/11 Commission concluded that to carry out catastrophic acts of terror, terrorist groups require sanctuaries that provide them with "time, space, and ability to perform competent planning and staff work" as well as "opportunities and space to recruit, train, and select operatives with the needed skills and dedication." Al Qaeda gained this in Pakistan with the signing of the South Waziristan accord, and later the North Waziristan accord, which signaled Musharraf's military defeat in the campaign directed at the tribal areas. The accords provided that Pakistan's military would not carry out air or ground strikes in the tribal areas, and included a pledge that Islamabad would disband its human intelligence network there. Three similar accords have since been signed: with the Bajaur region in March 2007, two months later with Swat, and finally with the Mohmand agency in August 2007. With these agreements in place, the United States has seen an influx of al Qaeda operatives and money into the tribal regions. Video taken in a Pakistani training camp last summer shows a graduation ceremony of about 300 recruits for suicide missions, some of whom are allegedly bound for the U.S. and Europe.

Compounding the problem of al Qaeda's Pakistan refuge, there are other areas where the group may gain further safe havens. One is Somalia, where most of the country was conquered by the al Qaeda-linked Islamic Courts Union in 2006. Though Ethiopian military intervention pushed that group back in 2007, the country now faces a potent Iraq-style insurgency spearheaded by extremist groups. And although the success of the U.S. troop surge over the past year has diminished the chances of an al Qaeda safe haven arising in Iraq, in the end that is a question of political will. If the next administration decides to quickly withdraw U.S. troops, that country could host additional safe havens.

Analysts declared al-Qaeda’s central leadership defeated before it had been dealt a death blow. Its regional nodes and ambitious newcomers stepped to the fore while the group's senior leadership fought to gain control of territory--thus helping to reinforce the idea that the senior leadership was marginalized and irrelevant. Even at the time, the fallacy of this view should have been apparent: As Peter Bergen noted in a New Republic article about al Qaeda's resurgence, "the existence of Al-Qaeda imitators does not prove the obsolescence of the real thing." Now, as al Qaeda's vitality approaches pre-9/11 levels, many analysts still do not have their eye on the central network.

With a safe haven in Pakistan--and perhaps soon in other territories--the senior leadership will likely play a greater role in future terror plots, while attempting to conceptualize and carry out an attack that will surpass 9/11. A strong central leadership makes the group more formidable and its attacks more deadly; dismissing the evidence that al Qaeda's leadership has regrouped will ultimately endanger U.S. security.