Few spectator sports rival the thrill of watching messengers being shot. Last week the editors of the august British Medical Journal were up before the firing squad over their decision to publish a study, part-funded by the tobacco industry, suggesting passive smoking was safe.

Could a shibboleth of public health be about to fall? Or had the evil tobacco industry infiltrated the pages of a respectable scientific publication with its campaign of misinformation?

No prizes for guessing what most people thought. The American Cancer Society, which had launched the huge study of 118,000 Californians in 1959, was the first to take aim, saying: "We are appalled that the tobacco industry has succeeded in giving visibility to a study with so many problems".

The British Medical Association, which owns the BMJ, was the next surprise critic, describing the study as "fundamentally flawed" and adding: "Most of the data has [sic] been around for decades but was judged by many expert groups to be inadequate to accurately measure passive smoking".

So had the BMJ been duped? Richard Smith, its editor and no stranger to controversy, offered a robust response. The paper had been thoroughly peer reviewed, the source of funding had been disclosed and it was accompanied by a critical commentary. "This is a big study with very complete follow-up about an important question. I take the view that not to publish is a form of scientific misconduct," Dr Smith said.

He was backed by Professor Martin Jarvis of the University of London, a leading expert on smoking research. "Science is science and one must not take the view that anything which has got any association with the industry is wrong," he said.

Anyone can level accusations of bias. In a press release I received last week, the pro-smoking group Forest says that the Californian study was funded by "anti-smoking largesse" - a levy on cigarette sales - until the plug was pulled in 1999.

"There is a strong suspicion that funding was stopped when the anti-smoking lobby saw the initial results and realised how damaging they might be to their claim that passive smoking harms non-smokers," Forest says.

So, bias cuts both ways. Is it more alarming that the study was funded by the tobacco industry - or not funded by the anti-smoking lobby? There is only one way to settle such disputes: publish and be damned. It may be uncomfortable, it may cause damage, but it is preferable to undermining public trust and denying readers the chance to make up their own minds.