Two recently released U.S. government reports are decidedly guarded in assessing the Pentagon’s fledgling national missile shield, with neither venturing much beyond the Missile Defense Agency (MDA) director’s remark last year that the system has a “better than zero” chance of knocking down an incoming warhead.

The reports are the latest indication of lower confidence in the effectiveness of the Ground Based Midcourse Defense System (GMD), whose initial interceptors have been deployed at Fort Greely in Alaska and Vandenberg Air Force Base in California. The system is more than a year behind schedule to go on alert.

The reports came from the Pentagon’s senior testing official and the non partisan Congressional Research Service. Their tone is a far cry from testimony by Edward “Pete” Aldridge — then the Pentagon’s top acquisition official — who told the Senate Armed Services Committee in March 2003 that the system had a 90-percent chance of destroying an incoming missile launched from North Korea.

Aldridge’s remarks stand in contrast to the opinion expressed by Air Force Lt. Gen. Henry “Trey” Obering, MDA’s director, who told reporters during a recent breakfast here that the system likely has a “better than zero chance” of intercepting missiles fired at the United States.

Obering said internal success-rate estimates for the GMD are classified. But he said his confidence in the system likely would rise as the MDA made improvements and increased its interceptor stockpile.

Steven Hildreth, a defense specialist with the Congressional Research Service, addressed both the national shield and systems meant to protect troops deployed around the world in a Jan. 18 report entitled “Kinetic Energy Kill for Ballistic Missile Defense: A Status Overview.” The service prepares reports for lawmakers and staffers and generally does not make them public, but Hildreth’s paper was obtained and posted online by the Federation of American Scientists, a non profit research group based in Washington.

Hildreth’s report says there is “insufficient empirical data to support a clear answer” on whether the GMD system, built by Chicago-based Boeing Co., can protect the United States from a missile attack. The report uses the term “ambiguous” to characterize the likelihood that similar technology can ever provide an effective defense.

Hildreth noted that the GMD program has roots in several MDA development efforts that have a cumulative record of seven intercepts in roughly 18 attempts — a success rate of about 39 percent. The Pentagon has yet to demonstrate a learning curve, such as a steadily improving success rate over time with the intercept tests, he said.

The Congressional Research Service report offered a more positive assessment of theater missile defense systems, particularly the Patriot and Aegis Ballistic Missile Defense System that evolved from existing air-defense programs.

In a separate report released to the media in January, David Duma, the Pentagon’s director of operational test and evaluation, was noncommittal in assessing the GMD system’s capabilities. “Developmental testing to date indicates that the GMD system may have some inherent defensive capability against a limited missile attack,” said Duma’s “FY 2005 Annual Report,” which evaluated the testing regimes for number of major Pentagon weapon acquisition programs.

There may be more to Duma’s GMD assessment than meets the eye, according to Victoria Samson, a research analyst with the Center for Defense Information here and a frequent critic of U.S. missile defense efforts. She said it can be difficult for Pentagon employees to truly speak their mind on missile defense, and suggested Duma’s characterization w as a diplomatic way of saying that the system would barely work.

Duma’s report says the MDA responded “appropriately” to test failures in late 2004 and early 2005 — in which GMD interceptor rockets failed to take off — by restructuring the program to focus on fundamental issues while gradually increasing the complexity of the tests . The MDA has yet to run tests that accurately mimic the conditions of a missile attack, in part due to the immaturity of some GMD system components, the report states.

Rick Lehner, an MDA spokesman, said the agency’s confidence in the GMD system has grown based on a successful booster flight test conducted in late 2005. That test did not include a target. The MDA is working to increase the realism of its tests this year through measures like launching interceptors from the operational GMD base at Vandenberg rather than the Kwajalein Atoll in the Pacific Ocean, Lehner said.