There's a lot of discussion about nuclear weapons timelines in Iran, and I thought it'd be interesting to lay out the most-comparible timeline to nuclear capability, that of Pakistan. This can hopefully serve as a factual anchor for our future discussions.
Obviously, Iran - assuming they got full cooperation from Pakistan's experts - could move faster. The interesting question is "how much faster?" given the technical issues involved in implementing both enrichment and weapons production.
I've based the timeline below on two sources: William Langweische's article on AQ Khan in the Atlantic, and the Nuclear Weapons Archive, a very useful site founded by Gary An, a student, and now operated by Carey Sublette.
Here's Pakistan's timeline (I've bolded the date that marks where in the process Iran is generally believed to be today):
1972 - PM Zulfikar Ali Bhutto starts Pakistan's nuclear weapons program
1974 - India detonates it's first nuclear weapon; AQ Khan, working in the Netherlands, meets Bhutto and begins assembling data on enrichment technology
1975 - AQ Khan moves from the Netherlands to Pakistan
1976 - AQ Khan founds Engineering Research Laboratories to build an enrichment facility
1978 - prototype and first enrichment
1981/2 - first weapons-grade uranium
1983 - 'inert test' of bomb design
1984 - production levels of 90% enriched uranium
1985/6 - weapons produced
So for Pakistan, it seems that it took 7 or 8 years to go from first enrichment to reasonably reliable - and hence usable - weapons, with few measurable waypoints along the way (I'm presuming that we are likely to have means sensitive enough to detect an inert test - which is a test of a weapon with unenriched uranium to make sure that the mechanics work).
Then the interesting questions become: "How far is Iran really down the road today?" and "How much faster than Pakistan can they get to the end?"