By Jonathan Marcus
Diplomatic correspondent, BBC News
Comments from John Negroponte, the Bush administration's most senior intelligence official, that Iran might have a nuclear weapon within the next 10 years, underline that the US is eager to keep up the pressure.
John Negroponte said Iran could have a bomb within 10 years
But just how accurate are such predictions?
Any assertion by an intelligence official about Iran's nuclear programme has to pass one very fundamental test.
If the combined effort of Washington's intelligence machinery was so wrong about Iraq's weapons programmes, then why should its assertions about Iran be seen as any more accurate?
It is a fair question and one that Mr Negroponte sought to answer by saying that some of the lessons of past failings had been learned.
There is, of course, a fundamental difference between Iran and Iraq.
Prior to its invasion of Kuwait, Iraq was, for many years, pursuing a nuclear weapons programme that largely eluded international inspectors.
Subsequently, the game of cat and mouse with the weapons inspectors continued, this time largely focused on chemical and biological programmes.
Western governments made their intelligence assessments to a large extent by extrapolating from the Baghdad regime's past behaviour.
And, in the event, these assessments turned out to be wrong.
With Iran though, there is no doubt about the nuclear activities that are under way.
Iran insists its nuclear work is purely for civilian purposes
What is at issue is their ultimate purpose - the uncertainty having been fostered by a whole range of areas where the Iranians have been less than forthcoming.
Much of the reliable information on Iran's activities comes from the routine work of International Atomic Energy Agency inspectors.
Their activities continue, though Iran has gone back on a pledge to allow them far greater access to its facilities.
Iran had voluntarily agreed to adhere to a more intrusive inspection regime - the so-called Additional Protocol - though it changed its mind and resumed its enrichment activities.
Additional intelligence - say from Iranian opposition groups - has sometimes also been helpful.
And then, of course, there is the pace of Iran's nuclear developments. If it wanted, despite its denials, to develop a nuclear weapon, how long might this take?
US estimates put this at anything from four to 10 years away.
The imprecision is a product, in part, of the limitations of intelligence gathering, but also because there are so many imponderables.
The rate of Iran's progress is hard to estimate.
Iran's recent and much-trumpeted enrichment experiments may well have been conducted with Chinese feed-stock rather than their own.
Western nations suspect Iran may be building a nuclear bomb
This suggests uncertainty about the quality of Iranian-processed uranium hexafluoride.
Will Tehran be able to overcome the obvious technical hurdles? Will they get any assistance from outside?
And then there is another question in relation to the timetable - just what is being measured?
What is the benchmark? Is it Iran having a fully-fledged nuclear bomb by 2010 or merely the capacity to manufacture sufficient fissile material for such a weapon?
Some would argue that in purely technical terms Iran is already well on the way to mastering the necessary procedures.
Nonetheless many analysts believe that the intelligence picture of Iran is patchy at best.
And it is as hard to determine who really is in control in Tehran, as it is to decide what the government's nuclear intentions really are.