A key US arms control official on Tuesday hailed a five-year-old US-led initiative aimed at halting trafficking in weapons of mass destruction, but offered little proof of the scheme's success.

The Proliferation Security Initiative was unveiled by President George W. Bush in Krakow, Poland on May 31, 2003, with a view to improving global coordination to intercept weapons shipments by rogue states and terrorist groups.

A year later, the White House declared the initiative a "great success" and said it played "a central role in our overall efforts to counter WMD (weapons of mass destruction) proliferation."

On Tuesday, when reporters asked for proof that PSI was as effective as claimed, under secretary of state for arms control and international security John Rood outlined "a number of successes" for the initiative, including international exercises and the burgeoning numbers of signatories to the US-led pact.

But he refused to go into detail about how many illicit arms shipment have been thwarted since PSI was launched.

"We have released some examples of successes but there are intelligence and other issues involved," he said.

"There are reasons why, when information has been clandestinely acquired, you want to protect that to the extent that you can from public disclosure.

"Depending on the circumstances of the interdiction, various people will know, or they may not know, exactly what led to it," he said.

He described PSI as having "grown to be recognized as one of the standards for non-proliferation behavior around the world," and said its aim "was first to build awareness and support among nations to stop proliferation-related shipments," he said.

"Metric (of success) number one: we have 90 countries participating in just five years," Rood said.

"Are those countries really committed? Are they working together more? I think there we have very good metrics as well," he said.

US National Security Adviser Stephen Hadley is due to hold a news conference on Wednesday about PSI, Rood said, urging reporters not to "measure PSI's success from the number of scalps."