Iran's Nuclear Program From Natanz in 2002 to the JCPOA's End, a Timeline

Governments that run peaceful nuclear programs usually announce them with a ribbon-cutting. Iran's biggest facilities were announced by other people, usually people who disliked Tehran. That is the pattern this timeline documents, and it is the pattern that matters most to anyone trying to judge what the regime means to do.

Let me state the claim so you can test it. Iran's nuclear program, from 2002 on, shows a repeated sequence: a facility is built in secret, an outside party exposes it, Iran admits what can no longer be denied, and negotiators then treat the admission as a starting point. That is a claim about a documented pattern. It is not a claim about what any official intends, and I will keep fact and judgment apart as we go.

Some terms. Enrichment is the process of raising the share of the fissile uranium isotope, uranium-235, in a sample. Low-enriched uranium, a few percent, fuels power reactors. Weapons-grade uranium is much higher, usually 90 percent. The IAEA is the International Atomic Energy Agency, the United Nations body that inspects nuclear facilities. Safeguards are the agreements that let it do so. The NPT is the Nuclear Non-Proliferation Treaty, which Iran signed and ratified, and which permits civilian nuclear energy while barring weapons development.

2002 to 2003: the revelation

August 14, 2002. At a press conference in Washington, an Iranian opposition group, the National Council of Resistance of Iran, disclosed the existence of two undeclared nuclear sites: an enrichment plant at Natanz and a heavy water facility at Arak. Why it mattered: the NPT safeguards system relies on states declaring their facilities. Neither site had been declared. The information came from an opposition group with its own agenda, which means we should treat its motives with caution, and then note that inspectors confirmed the sites existed.

February 2003. IAEA Director General Mohamed ElBaradei visited Natanz and saw an advanced facility under construction, including centrifuge halls, with plans for much larger work. Iran said the program was entirely peaceful and that it had a right to enrich under the treaty. The right to peaceful nuclear energy is real under the NPT. The failure to declare is the problem.

June and November 2003. IAEA reports said Iran had failed over a period of many years to report nuclear material, facilities and activities that it was required to report. In November 2003 the Board of Governors adopted a resolution that deplored these failures, though it stopped short of declaring non-compliance.

October 2003. Iran signed an agreement with Britain, France and Germany, known as the E3, in which Tehran agreed to suspend enrichment and sign the Additional Protocol, a stronger inspection regime. Why it mattered: this was the first serious diplomatic channel, and it set the template for later negotiations. A suspension, remember, is a pause, not a dismantling.

2004 to 2009: the Security Council

2005. After the election of Mahmoud Ahmadinejad as president, Iran resumed uranium conversion at its Isfahan facility, ending the suspension it had accepted with the Europeans.

February 2006. The IAEA Board of Governors voted to report Iran's file to the United Nations Security Council. Why it mattered: this moved the question from a technical agency to the body with power to impose sanctions.

2006 to 2010. The Security Council passed a series of resolutions. Resolution 1696 in July 2006 demanded that Iran suspend enrichment. Resolution 1737 in December 2006 imposed sanctions on nuclear and missile-related transfers, and Resolutions 1747 in 2007 and 1803 in 2008 widened them. Resolution 1929, on June 9, 2010, was the broadest yet, adding an arms embargo and restrictions on financial dealings. Iran rejected the demand to suspend. Note what these resolutions mean in plain terms: the world's highest security body spoke repeatedly and Tehran did not comply.

December 2007. A U.S. National Intelligence Estimate judged with high confidence that Iran had halted its nuclear weapons program in the fall of 2003, while continuing enrichment, which keeps the ability to build a weapon in reserve. The finding was controversial, and it is cited by both sides. Defenders of the engagement approach say it shows no weapons program was in progress. Skeptics answer that the same finding confirms there had been a weapons program to halt, and that it left enrichment capacity untouched.

September 25, 2009. President Obama, with the leaders of Britain and France, stood in Pittsburgh and announced the existence of a second enrichment facility, dug into a mountain near the city of Qom, called Fordow. Iran had told the IAEA about it only days before, by letter, after learning that Western intelligence services knew. Why it mattered: this is the third time in the story that an undeclared facility came to light under pressure. A site built inside a mountain is not built to produce electricity for a city. It is built to survive an attack.

2013 to 2015: the deal

November 24, 2013. In Geneva, Iran and the six negotiating powers, the United States, Britain, France, Germany, Russia and China, agreed an interim deal, the Joint Plan of Action. Iran agreed to limit certain enrichment and to allow more inspections, and received limited sanctions relief.

July 14, 2015. After a long negotiation, Iran and the six powers announced the Joint Comprehensive Plan of Action, the JCPOA, in Vienna. Its main terms:

• Iran would limit enrichment to 3.67 percent.
• Iran would keep a stockpile of no more than 300 kilograms of that material.
• Iran would operate no more than about 5,060 older-model centrifuges at Natanz for ten years.
• Fordow would stop enriching and be converted for other uses, for fifteen years.
• The Arak reactor would be redesigned to reduce its plutonium output.
• The IAEA would receive expanded access, though with a dispute process that could delay access to undeclared sites.
• In exchange, American, European and United Nations sanctions tied to the nuclear program would be lifted.

Why it mattered: it was the largest diplomatic agreement on the question, and the argument over it split the country. Supporters said it verifiably blocked the paths to a bomb for a decade or more. Critics said it left the enrichment infrastructure in place, expired on a schedule, and did nothing about ballistic missiles or Iran's support for armed groups across the region.

The Obama administration did not submit the deal as a treaty. Instead, it relied on Congress's review under the Iran Nuclear Agreement Review Act, which gave Congress a review period and a chance to pass a resolution of disapproval, but never put the deal to the two-thirds Senate vote that a treaty requires. That is a constitutional question worth thinking about. Article II, Section 2 of the Constitution says treaties need two thirds of the Senators present. An agreement of this weight that bypasses it is an instructive case of how modern presidents treat the treaty power, a subject the Founding documents on the Constitution page let you read for yourself.

January 16, 2016. "Implementation Day." The IAEA certified that Iran had completed its main steps, and sanctions relief took effect.

2018 to 2021: withdrawal and escalation

April 30, 2018. Israel's prime minister presented what he said was a large archive of Iranian documents about the country's earlier weapons work, which he said Israel had obtained from a warehouse in Tehran. He said it showed a program that Iran has always denied having. Critics noted that the material concerned the years before the deal and did not show a breach of it, and the IAEA said it would review the material rigorously and would not take it at face value. Evaluating it is a job for experts.

May 8, 2018. President Trump announced that the United States would withdraw from the JCPOA and restore sanctions. His reasons: the deal's sunsets, its failure to address missiles, and the claim that Iran had lied about its past. The Europeans, Russia and China stayed in the agreement. Why it mattered: the pact survived on paper, but the economic relief Iran expected shrank as American sanctions bit.

2019. Beginning in the summer of 2019, Iran began to exceed the deal's limits step by step. It went past the 300-kilogram stockpile cap and the 3.67 percent enrichment ceiling, enriching to about 4.5 percent, and in November 2019 it resumed enrichment at Fordow. Iran said these were responses to the American withdrawal and to Europe's failure to deliver economic benefits. Both sides have a story about who broke first. The fact is that each step moved Iran closer to a weapons-capable stockpile.

2021. Iran began enriching up to 20 percent in January and to 60 percent in April. Sixty percent has no ordinary civilian use that requires it, and it sits a short technical step from weapons-grade.

2022 to 2025: the end of the agreement

Negotiations in Vienna to revive the deal went on through 2021 and 2022 without result. Meanwhile the IAEA reported repeatedly that it could not answer basic questions about Iran's stockpile and its past undeclared activities, and that its monitoring had been reduced.

August 28, 2025. Britain, France and Germany triggered the "snapback" mechanism built into the agreement and into United Nations Security Council Resolution 2231, which let a participant restore the earlier UN sanctions if it judged Iran to be in significant non-compliance. UN sanctions took effect again at the end of September 2025.

October 18, 2025. The date the JCPOA was scheduled to terminate, ten years after its adoption. Why it mattered: the sunsets that critics had warned about arrived on the calendar. Whatever you think of the pact, the timetable it built was always going to run out.

What the pattern shows

Now the argument, with numbered premises.

1. Iran built nuclear facilities without declaring them. Natanz and Arak, disclosed in 2002, and Fordow, disclosed in 2009 under pressure, were all undeclared when they were built. That is a matter of record.
2. The IAEA found years of failures to report. In 2003 it documented breaches of Iran's safeguards obligations.
3. The Security Council passed five enforcement resolutions in four years and Iran did not comply. The resolutions came between 2006 and 2010.
4. Every agreement was a pause and not a rollback. The suspension of 2003, the 2013 interim deal and the JCPOA each limited enrichment for a period without eliminating the capability.
5. Capability is the issue, not intent. A state that can enrich to 60 percent has covered most of the technical distance to 90 percent. What remains is a short technical step, not years of research.

That leaves the steelman. The best case for the engagement approach says the JCPOA did what it was designed to do: it froze the program, put inspectors in place and extended the time Iran would need to build a bomb from months to about a year. America's withdrawal, on this view, threw away those limits and Iran then escalated. A serious critic of the withdrawal can say the deal was imperfect but better than no deal.

The answer: grant the technical point. For the years it was fully observed, the JCPOA did cap the stockpile and the centrifuges, and I do not dispute that. But the capping was temporary by design, which is the heart of the criticism. A deal whose key limits expire on a schedule gives the other side a date to wait for. And the deal left untouched the pattern in the early entries of this timeline: a state that hides, is caught, admits and bargains. Whatever one thinks of how the withdrawal was handled, a pact that depends on the good faith of a party with that record needs verification that does not depend on that party's cooperation. The inspection regime was the weakest part.

Where this fits

The Iranian program is one thread in a larger story of how America and its allies deal with regimes that seek the most dangerous weapons. Israel's decision to strike Iraq's reactor in 1981, told in Israel Bombed Osirak in 1981 and the World Later Thanked It in Private, is the precedent that shaped Israeli thinking about Iran. The lesson that force has to be credible if diplomacy is to work is the theme of Deterrence Fails When an Enemy Thinks You Will Not Pay the Price. The arms-control record on when treaties succeed, covered in The INF Treaty Proved Arms Control Works Only When You Hold Cards, is a useful comparison. And the longer record of Iranian behavior, from the 1983 Beirut bombing in Beirut 1983 Was the First Lesson in How Iran Fights Through Proxies to the naval clash described in Operation Praying Mantis and the Day the Navy Answered Iran in 1988, shows the regime that Western negotiators have been dealing with.

A wry note for the reader who follows the news cycle: every few years a headline announces that Iran is "weeks away" from something, and a rival headline announces that the first headline was exaggerated. Both can be true, which is exactly why the dated record is worth more than the cable chatter. Dates do not have a side.

The case in brief

1. Iran's biggest nuclear sites at Natanz, Arak and Fordow were exposed by outsiders, in 2002 and 2009, and were not declared by Tehran.
2. The IAEA found in 2003 that Iran had failed to report nuclear activities, and the Security Council passed repeated resolutions from 2006 that Iran did not obey.
3. The JCPOA of July 14, 2015, capped enrichment and the stockpile for a period, and its critics and defenders both have a real case.
4. After the 2018 U.S. withdrawal, Iran moved step by step past the limits, reaching 60 percent enrichment in 2021, and the agreement reached its scheduled end in October 2025.
5. The lasting lesson is that verification, not trust, has to carry any agreement with a regime whose record is concealment.

What you can do: read the IAEA's public Board of Governors reports on Iran, which are posted on the agency's website, and ask your representative whether Congress has been shown the full verification terms of any agreement.