It is 41 years since Alec Jenner was sitting in a Sheffield hospital canteen, reading in the Daily Mirror how a circus trainer was using a drug to sedate his tigers. "Hey, this could work on the people of Sheffield," he joked to his colleagues.

But the ambitious young psychiatrist saw in the story more than material for a wisecrack; he'd spotted an opportunity to further his career. He wrote to Swiss pharmaceutical company Hoffmann La Roche, suggesting the drug could relieve anxiety in humans and offering to carry out trials on patients. The company agreed and, after Jenner and his colleagues' research yielded positive results, Roche developed Valium.

The drug was launched on to the global market in 1963. Over the next 40 years, Valium and similar tranquillisers called benzodiazepines became legendary for their immediate relief of anxiety and depression for millions across the world.

Jenner, now 72 and retired, should therefore perhaps be glowing with pride in his farmhouse near Sheffield, after playing his part in the creation of the drug, known affectionately as "mother's little helper". But, over the past 20 years, he has witnessed Roche and other benzodiazepine manufacturers feeling the wrath of thousands of users, complaining of addiction and brain damage they claim has been caused by long-term use of the drugs.

In 1986, 17,000 people brought a group action against benzodiazepine makers Wyeth, Roche and Upjohn. Although the action collapsed in 1994, when the legal aid board stopped paying legal fees that had already reached £35m, the sense of injustice felt by the victims is as strong as ever. At a "beat the benzos" conference earlier this month, calls were renewed for compensation for benzodiazepine victims and a full inquiry into how the drugs were prescribed, despite their having been listed as drugs of "dependence" by the World Health Organisation and the British Medical Journal as early as the 60s.

In fact, it took until 1988 - two years after the group action launch - for the committee of safety of medicines to recognise benzodiazepine dependence and to introduce guidelines specifying that the drugs should not be prescribed for more than four weeks.

Jenner admits that when he carried out his early trials, he did not even consider the possibility of dependency problems - for which he is left feeling "extremely naïve". He says: "When I was researching the drug, we just thought of the advances this new drug could make. In the late 50s, the main treatment for anxiety was the barbiturate drugs, which were a lot more dangerous. When I was an ambitious doctor, I did not even think of Valium's addictive potential - although I now wish I had."

Jenner admits that research on long-term effects of medication is not a cosy fit with the profit-making plans of pharmaceutical companies. "Long-term trials are very expensive and difficult to arrange," he says. "You cannot take this out of the equation in assessing what happened with benzodiazepines.

"And I would have been surprised if Roche had asked me to do long-term trials of Valium. The company would have been anxious to get it on the market. This is where you can be critical of the drug companies - but it is the system in which they work."

After his early pioneering trials, Jenner did little further work for Roche. But, in 1990, he received a letter from the company's solicitors asking him to give his scientific opinion on the files of dozens of the group action Valium users complaining of addiction and memory loss, epileptic fits, vision problems, mood swings and extensive cognitive deficits. His assessment was that, generically, the users' problems were as likely to be due to "addictive personalities" and "neurotic" or "pre-morbid" diagnoses as Valium addiction and damage.

Roche used Jenner's analysis in its defence. A company spokesman says its benzodiazepines have, since launch, been fully researched and kept under constant review. "They have come to play an essential role in modern clinical practice, not only in the treatment of sleep and anxiety disorders but in a wide variety of psychiatric and medical uses." The litigation had been "misguided and bound to fail".

Jenner says now: "A lot of files I examined were those of people who have what I call 'problems of living' and who are searching for explanations of why they have such problems. I have sympathy for the people involved, some of whom may have a case. But the issue for me was to demonstrate scientifically whether their complaints were because of the effects of Valium. This was not the case. And, in legal cases, the benefit of the doubt is always with the accused.

"Having said this, much of the research is contradictory and we are not certain of very much in psychiatry. Allegations of damage could turn out to be true - in the same way as we did not originally think of the possibilities of addiction but now recognise this to be so."

Jenner agreed to talk to the Guardian because he believes in open discussion about psychiatry and its drugs. His philosophy was demonstrated by him becoming founding editor in 1986 of Asylum - A Magazine for Democratic Psychiatry, which questions traditional psychiatric practice and continues the debate about benzodiazepines - particularly raising concerns about the high levels of the drugs given to patients in secure psychiatric units.

And despite Jenner's view that the drug companies have no case to answer, he still recognises some validity in calls for a public inquiry. "Such an inquiry might find something I have missed," he says. "It might satisfy people who think that something has been hidden in the whole story about benzodiazepines."

• Asylum is published quarterly by Handsell Publishing (01452-380319)