Cadbury Beverages, now Cadbury Schweppes, privately withdrew its Diet Orange Crush drink in five regions in the US in 1990 because tests found some products with benzene several times above the acceptable limit. Benzene is a cancerous chemical.
The revelation, confirmed by newly released official documents passed to BeverageDaily.com, raises new questions about the seriousness of the benzene in soft drinks problem.
Cadbury's Diet Crush contained 36 and 52 parts per billion (ppb) benzene in two samples taken by the Food and Drug Administration (FDA) late 1990. Another preliminary sample showed 130ppb benzene, although others were much lower.
The industry limit is historically 10ppb, although no official maximum exists.
Perrier spring water was publicly recalled earlier the same year at less than 20ppb benzene. The brand's sales have never properly recovered in North America.
Perrier's problems emanated from a one-off processing glitch, but Cadbury's benzene was traced to a reaction between two common ingredients in the drink - something that was still producing benzene in some other firms' drinks this year.
Several drinks have been reformulated this year since the FDA revealed it had found samples with benzene above the five ppb limit for US drinking water. Another four drinks were recalled in the UK this spring as news of the problem spread.
Both the FDA and soft drinks firms have always said there was no risk to health from benzene levels found in beverages.
Lawyers investigating the issue praised Cadbury for "doing the right thing" by informing the FDA of its benzene problem in 1990 and immediately reformulating Diet Orange Crush.
But the problem with the two ingredients - sodium benzoates and citric or ascorbic acid (vitamin C) - was never made public.
And their wide use together in drinks formulas raises questions about which other brands may have contained excessive benzene in 1990.
"Levels of benzene formed are highly variable with treatment, but ranged from four ppb to several hundred ppb," an FDA memo dated January 1991 says. "Even sitting at room temperature in the dark yield [sic] low benzene levels after one or more days." Heat and light exposure rapidly increased benzene formation.
Another memo, dated February, says benzene levels found in Diet Orange Crush posed no acute hazard to health, however. "Consumption of the contaminated beverage for relatively short periods of time would not be expected to increase detectable risk of cancer," it added.
FDA made a deal with soft drinks firms in 1990 for them to "get the word out and reformulate", instead of taking the matter public, one agency scientist there at the time told BeverageDaily.com.
But isolated drinks samples continued to show elevated benzene levels in later FDA Total Diet Surveys.
It is "embarrassing" the FDA failed to eradicate benzene residues from all drinks, a senior, ex-FDA enforcement official said, speaking on condition of anonymity.
"Big companies are very powerful. If you're a regulator with a tight budget, it could have been one of those closets with skeletons in that you don't want to open."
An industry-wide guidance document on limiting benzene in drinks, the first to be compiled, was published by the International Council of Beverages Associations after the issue went public this year.
Some say the public had a right to know what was happening in 1990 and that Cadbury had pulled one of its drinks.
"It's outrageous that when a known carcinogen such as benzene was found in popular USA soft drinks, neither the beverage industry or FDA notified consumers," said an ex-Cadbury scientist, who approached BeverageDaily.com alongside lawyer Ross Getman about the benzene in soft drinks problem.