On a night of sleet, a tailings pond overtopped an earth dam. One hundred thousand cubic metres of cyanide water set off down five rivers — and four weeks later, the plume was still measurable in the Danube Delta, 2,000 kilometres away.
2,000 km of river touched
3 countries on the plume's path
1 judgment in Strasbourg
The buoy on the bank: the plume showed itself through dead fish.
At ten in the evening, a worker on the Aurul site watched water cross the dam crest and raised the alarm. Rain was falling on 60–70 centimetres of snow, the temperature had risen past freezing, and the tailings pond — 93 hectares of cyanide slurry — no longer fitted inside the earth dam holding it.
At eleven at night, the company told the Baia Mare Environmental Protection Agency of "incidents" at its installations. The field inspection settled it: the dam had overflowed and washed away along some 25 metres, down to 2.5 metres deep. About 100,000 cubic metres of water carrying free cyanide and cyanide complexes left the pond, covered the surrounding land and reached the Lăpuș river.
The pond held the poison of a gold business: the plant reprocessed old tailings to draw about 1.6 tons of gold and 9 tons of silver a year from them, and the cyanide solution washing the ore had reached some 400 milligrams per litre — around 4,000 times the Hungarian drinking-water norm of 0.1 milligrams per litre.
The leak ran until 2 February at half past one in the morning, when the breach was sealed. Meanwhile the company dosed it with sodium hypochlorite, and the remaining water was pumped into the neighbouring Remin pond. Decontamination covered about 14 hectares. The worst-hit village was Bozânta Mare, straight in the flood's path: days later, experts found eight wells contaminated with cyanides there.
What left the pond
100,000
cubic metres of cyanide water — 40 Olympic pools, discharged in a single night. The UN mission's estimate: between 50 and 100 tons of cyanide reached the rivers.
Source: UNEP/OCHA report, March 2000 · pools at 50 × 25 × 2 m
On 2 February, Satu Mare logged the first dead-fish report. The plume had left the Săsar for the Lăpuș and the Lăpuș for the Somes, and at Csenger, on the Hungarian border, the instruments caught copper at 18 milligrams per litre — 180 times the health limit.
From there the poison kept the river's pace: about 2 kilometres per hour on the Tisza, 14 days to Yugoslavia, then 1,200 more kilometres down the Danube to the Black Sea. Hungary briefly closed the Kisköre dam to raise the water level and dilute the plume; the Szolnok waterworks stopped while the pollution passed, and Belgrade closed its Danube intake as a precaution.
Each stop below is a measured peak with its day. The Tisza values come from the Hungarian monitoring of those days, later compiled from terra.hu data; the pond and delta values come from the UN reports. The UN mission honestly notes that Romanian and Hungarian readings do not match everywhere — sampling places and hours differed.
Baia Mare, the Aurul pond≈400 mg/L
The 25-metre breach. Total cyanide in the pond had reached about 400 milligrams per litre, some 120 of it free.
Csenger, the Romanian–Hungarian borderCu 18 mg/L
Copper peaks at 18 milligrams per litre — 180 times the health limit. The plume's first heavy mark.
Satu Marefirst dead fish
The first dead-fish report on the Somes. At Baia Mare, the pond leak is stopped at half past one in the morning.
Tiszafüred3.4 mg/L
The upper-Tisza peak. Below Kisköre, the closed dam holds values under 3 milligrams per litre.
Szolnok2.3 mg/L
The town's waterworks stops for the plume; treated water stays under the Hungarian 0.1 mg/L norm.
Csongrád and Mindszent2.9 / 2.0 mg/L
The Csongrád peak at midnight; at Mindszent, eight in the evening, the plume passes at 2.0 milligrams per litre.
Szeged1.7 mg/L
At the last large city on the Tisza, the plume passes in the morning. At midnight, at Tiszasziget, it still reads 1.49.
The Serbian Tiszathe watch on the bank
The plume enters Yugoslavia. Serbia bans fishing and the fish trade; dead fish are collected and buried in sanitary pits.
The plume leaves Hungarybelow lethal levels
The contamination exits Hungary and runs down the Danube toward Belgrade, ever more diluted.
Danube Delta, Cheatal-Izmail0.058 mg/L
The UN team catches the plume in the delta: 0.058 milligrams per litre at most, above the European 0.025 criterion. Four weeks and 2,000 kilometres from the breach.
The plume's schematic road, from pond to sea. Dots mark measured peaks; colours mark dilution — from the pond's red to the delta's green.
over 10 mg/L
2–10 mg/L
1–2 mg/L
under 0.1 mg/L
Schematic map, not to scale: distances are compressed, while the order of places and the values are the measured ones. Tisza values: Hungarian monitoring of February 2000, compiled from terra.hu data. The UN mission warns that Romanian and Hungarian readings differ in places.
How far the plume went
2,000
kilometres of Danube basin touched — 800 to Yugoslavia in 14 days, plus 1,200 more on the Danube. The plume advanced at about 2 kilometres per hour: a strolling pace.
Source: UNEP/OCHA report, March 2000 · measured speeds 2.1–2.9 km/h
On the night from Friday to Saturday, 11 to 12 February, the poison wave reached Yugoslavia. The people on watch that night stood aghast at dead fish floating on the river's surface, everyone mentioning the smell of bitter almonds in the air. In the morning, dead birds could be seen too, which had drunk the river water or eaten the dead fish.
The scene, reported by the Belgrade weekly Vreme, had repeated itself over hundreds of kilometres in Hungary: pictures of enormous quantities of fish floating in agony went around the world. Phytoplankton and zooplankton fell to zero as the plume passed, and fish died in the plume or right after it — the UN mission's finding, traced from Romania to the delta.
How much fish died? The Hungarian authorities estimated over a thousand tons; the Romanian authorities reported a very small amount; Yugoslavia spoke of a large amount on its Tisza stretch. Greenpeace, five years after the accident, wrote of more than 1,400 tons and several hundred Tisza fishermen left without a living. No major fish kills were reported from the Danube: dilution had done its work.
The heavy word: "Chernobyl"
The Chernobyl comparison was uttered by name, by named people. We print it as spoken, with its authors beside it — then set the only measure that matters next to it: the measured numbers.
"The biggest environmental catastrophe since Chernobyl."
Zoltan Illes, head of the Hungarian parliament's environment committee, to the Associated Press, February 2000. He added that the heavy metals would "poison the whole food chain".
"The worst ecological disaster in central Europe since Chernobyl in 1986."
Hungarian politicians, quoted by Aaron Schwabach in the Virginia Environmental Law Journal, 2000.
"It would not be an exaggeration to put what has happened here in the past few weeks on a par with Chernobyl."
Jan Korabov, WWF spokesman, quoted in the same study.
The measure behind the words
The plume stayed measurable for four weeks, over 2,000 kilometres, to the delta.
In the delta: 0.058 mg/L — more than twice the European 0.025 mg/L criterion.
On the Tisza: plankton fell to zero; fish died by the hundred tons.
Nobody died and nobody fell gravely ill — the waters lay frozen for 200 kilometres, and the waterworks closed their intakes in time.
The dead fish of the Tisza
>1,000
tons, on the Hungarian authorities' estimate — over 50 truckloads of dead fish, gathered by volunteers to keep it out of the food chain. Greenpeace speaks of more than 1,400 tons.
Source: UNEP/OCHA report (Hungarian estimate) · Greenpeace, the 5-year account
On 27 January 2009, nine years after the night of the dam, the European Court of Human Rights read out a judgment in Strasbourg with Romania's name in the title. The Tătar family of Baia Mare had complained that the gold mine near their home endangered their lives — and that the Romanian state had neither assessed the risks nor informed the public.
The Court sided with the applicants: Romania had breached its positive obligations under Article 8 — the right to private and family life. It noted that the authorities had not stopped the mine's activity even after the accident, and invoked the precautionary principle explicitly. But since the causal link between the pollution and the deterioration of the son's health was not proven, it awarded no damages — with two judges dissenting, arguing that absolute causality is almost impossible to prove in toxic exposures.
Judgment · Strasbourg
Tătar v. Romania, application 67021/01
Verdict: violation of Article 8; no damages.
The only final court judgment tied to the accident. The Romanian state neither assessed the risks of the cyanide method nor published the studies — before or after the accident.
Why it matters
The judgment of 27 January 2009 ranks among the Court's landmark environmental decisions: it ties industrial pollution to private life and asks states to apply the precautionary principle rather than wait for proof of harm.
Announcement · Baia Mare
The environment agency explores legal action
Status: an intention on the UN record.
On 2 February 2000, while the first dead fish were reported at Satu Mare, the Baia Mare Environmental Protection Agency was exploring legal action against the Aurul company.
Announcement · Belgrade
Serbia says it will sue
Status: a statement from February 2000.
Serbian officials said they would sue those responsible in an international court, after the plume reached the Yugoslav Danube leaving dead fish behind.
Position · the company
Esmeralda: "not proven that we are the source"
Status: a public denial.
The Australian company Esmeralda, Aurul's majority shareholder (51%, the rest with the Romanian state), denied it through president Bret Montgomery on CNN: "It has not been proven that we are the source of the pollution in the Tisza" — always speaking of an "ecological incident".
A second breach, in March 2000 at Baia Borșa (the Novat dam, about 20,000 tons of heavy-metal sludge), struck rivers the cyanide plume had spared. It is a separate accident with a different pollutant — outside this account.
The year is 2000, the phone rings from upstream: the plume has passed them. Five Tisza and delta towns await your call — on which day does the plume reach each? Choose, then read what the instruments caught.
The watch on the Tisza. Five calls to make. Every right answer keeps a town's intakes closed in time.
Every number carries its source beside it. The estimates of the total cyanide differ — 50–100 tons from the UN mission, up to 120 from the international task force, 130–175 tons in a Hungarian mass balance — and appear separately, each with its author. Romanian and Hungarian concentration readings differ in places; the UN mission says it cannot fully explain why. The title is an editorial image: no document records a depth of thirty centimetres — the measured dimensions of the night are the ones in the text: the 25-metre breach, the 60–70 centimetres of snow, the 100,000 cubic metres.
The equivalents help feel the scale: an Olympic pool is 50 × 25 × 2 metres, or 2,500 cubic metres; 100,000 divided by 2,500 makes 40 pools. A truck carries about 20 tons; 1,000 tons make over 50 truckloads.
The night of the breach, the plume's road and the fish account come from the UNEP/OCHA assessment mission report (Geneva, March 2000) and the Baia Mare International Task Force report. The Tisza concentrations come from the Hungarian monitoring of February 2000. Quotations belong to the authors named beside them. The data follow the editions read on 20 September 2026.