How did a treaty agreed in 1987 and eventually ratified by every nation shrink a hole in the atmosphere the size of North America back toward 1980 levels?

The answer is the Montreal Protocol, the 1987 agreement that required countries to phase out chlorofluorocarbons and other ozone-depleting substances, tightened those controls as the science sharpened, and helped developing countries pay for the transition. Nearly 99 percent of banned ozone-depleting substances have now been phased out, according to the UN-backed 2022 scientific assessment. The latest complete Antarctic season, in 2025, produced the fifth-smallest average ozone hole since 1992, while a return to 1980 ozone levels over Antarctica is projected for around 2066 if current policies remain in place.

The hole has not vanished, and it does not shrink neatly every year. The treaty worked by cutting off the source chemicals, building a system that could become stricter over time, and waiting while chlorine already released into the atmosphere slowly declined.

A discovery in the Antarctic spring

In May 1985, British Antarctic Survey scientists Joe Farman, Brian Gardiner, and Jonathan Shanklin published a short paper in Nature. Their long-running measurements at Halley station showed that springtime ozone above Antarctica had fallen with extraordinary speed, a result documented in their original 1985 study.

The ground record established that this was not a bad instrument or a single strange season. Satellite observations then showed that the depleted region stretched across a continental area, turning an invisible chemical change into one of the most recognisable images in environmental history.

The underlying warning was already eleven years old. In 1974, Mario Molina and F. Sherwood Rowland had argued in a landmark Nature paper that stable chlorofluorocarbons could survive in the lower atmosphere, drift upward, and be broken apart by ultraviolet light. The chlorine released there could destroy ozone catalytically, then continue attacking more ozone molecules.

Antarctica supplied the conditions that made the chemistry explosive. During the dark polar winter, extremely cold air is trapped inside the polar vortex and polar stratospheric clouds form. Their particles convert relatively inactive chlorine compounds into reactive forms; when sunlight returns in spring, chlorine and bromine begin destroying ozone at great speed, as NASA explains.

A treaty built to tighten

The Vienna Convention of 1985 created a framework for research and cooperation but imposed no binding production cuts. The Montreal Protocol was adopted on September 16, 1987, entered into force in 1989, and set stepwise controls on the production and consumption of ozone-depleting substances, with different timetables for developed and developing countries. UNEP’s official overview also sets out the treaty’s trade, reporting, and licensing obligations.

The original deal required industrialised countries to reduce production and consumption of the main CFCs to 50 percent of 1986 levels by 1998-99. That was not the final destination. Later meetings accelerated the schedules, added more chemicals, and converted what began as a partial reduction plan into a near-complete phase-out.

It also did not begin with nearly every nation signing on the same day. Participation widened over more than two decades, and the Vienna Convention and Montreal Protocol achieved universal ratification on September 16, 2009, according to the United Nations.

That sequence matters because the protocol was not a single promise left untouched after 1987. It required regular scientific assessments, data reporting, national controls on imports and exports, and repeated decisions by the parties. The agreement could respond when measurements showed that the first targets were not strong enough.

The deal worked because the transition was financed

The treaty gave developing countries longer timetables, but delay alone would not have replaced refrigeration systems, redesigned aerosol production, or trained technicians. In 1991, the parties established the Multilateral Fund to support developing countries with project finance, technical guidance, and capacity building.

That changed the practical question. Governments were no longer being asked simply to ban useful chemicals and absorb the cost. They had a mechanism for converting factories, changing equipment, managing old chemical stocks, and meeting obligations on a schedule their economies could carry.

Industry also received something it could plan around: a global timetable. Once the direction of regulation became clear, manufacturers could invest in substitutes knowing that competitors would face the same transition. The protocol did not depend on a single replacement appearing overnight; it created a sequence of technologies and progressively tighter controls.

Hydrochlorofluorocarbons, or HCFCs, became an interim substitute because they damaged the ozone layer less than the CFCs they replaced, although they still contained chlorine. Hydrofluorocarbons, or HFCs, removed the ozone problem but introduced another one because many are powerful greenhouse gases. The continuing effort to reduce both refrigerant emissions and energy use now reaches beyond chemistry, including cooling systems that abandon the conventional compressor.

In 2016, countries adopted the Kigali Amendment to phase down HFCs under the same treaty system. Other approaches are also moving into buildings, including refrigerant-free cooling technology being developed in France. The replacement problem did not end with CFCs, but the Montreal framework proved capable of taking on the next chemical generation.

Why recovery takes generations

Stopping new emissions could not remove the gases already in the atmosphere. Many CFC molecules persist for decades, and some for well over a century, before they are destroyed. The chlorine they eventually release can remain active in the stratosphere for years.

That is why the Antarctic ozone hole still opens each southern spring. The polar vortex, extreme cold, cloud particles, and returning sunlight are recurring features of the atmosphere. What is changing is the quantity of human-made chlorine and bromine available to drive the reactions.

Weather can still overpower the long-term signal in an individual year. A colder, more stable polar vortex can support a larger or longer-lasting hole, while warmer stratospheric conditions can produce a smaller one. Volcanic particles and smoke reaching the stratosphere can also disturb the chemistry, so a single season is not a clean verdict on recovery.

The scientific assessment therefore measures trends across years and compares them with chemistry-climate models. Under current policies, total ozone is expected to return to 1980 values around 2040 for most of the world, around 2045 over the Arctic, and around 2066 over Antarctica. The later Antarctic date reflects both the depth of the damage and the slow atmospheric lifetime of the substances that caused it.

The climate dividend and the latest measurements

CFCs and several other controlled substances are also strong greenhouse gases. Their phase-out reduced warming as well as ozone loss, even though climate protection was not the treaty’s original purpose. The benefit came from preventing the continued growth of gases that trap far more heat per molecule than carbon dioxide.

A 2023 modelling study by Mark England and Lorenzo Polvani estimated that the Montreal Protocol has delayed the first nearly ice-free Arctic summer by as much as 15 years, depending on future emissions. The study in the Proceedings of the National Academy of Sciences attributed that delay to the greenhouse warming avoided by controlling ozone-depleting substances.

That does not make the ozone treaty a simple template for carbon dioxide. CFC production was concentrated in a limited set of industries, substitute chemicals could be developed, and the Multilateral Fund could target identifiable conversion costs. Fossil carbon runs through electricity, transport, buildings, food, and heavy industry, which makes the economic replacement problem far broader.

The atmosphere nevertheless shows a measurable result. NASA and NOAA reported that, from September 7 to October 13, the average area of the 2025 Antarctic ozone hole was 18.71 million square kilometres, the fifth-smallest average since 1992. Its one-day maximum reached 22.86 million square kilometres on September 9, still continental in scale but about 30 percent smaller than the largest hole observed in 2006, according to NASA’s season report.

Those numbers explain why both halves of the story can be true. The hole remains vast enough to cover an area approaching North America at its seasonal maximum, yet the multi-decade record is moving in the direction predicted when emissions of the source chemicals fall.

A recovery that still has to be watched

The protocol has already faced a serious compliance test. Atmospheric measurements detected an unexpected rise in CFC-11 emissions beginning in 2013, and later work attributed much of the increase to eastern mainland China. A 2021 Nature study found that emissions from the region had returned to pre-2013 levels by 2019 after a sharp decline.

The episode did not erase the recovery, but it showed why monitoring remains part of the mechanism. CFC-11 is colourless, its production can be hidden inside industrial supply chains, and its atmospheric lifetime makes a few years of unreported manufacture matter long after the factories stop.

The 2026 Antarctic ozone-hole season is only beginning as this article is written, so 2025 remains the latest complete season for comparison. No responsible account can promise that the coming hole will be smaller than the last one; winds and temperatures may push the annual number in either direction.

When sunlight returns over Antarctica, instruments will again watch ozone fall across the polar cap and then rebuild as the vortex breaks apart. The recovery will not appear as a smooth line on a chart. It will emerge across noisy seasons, while chlorine released by twentieth-century industry continues its slow retreat from the stratosphere.