The research councils are looking at what more they can do to support open access to research results after an independent study found that their current policies were having a "limited impact".
The councils have previously baulked at requiring all council-funded researchers to deposit papers in openly available repositories.
Instead they adopted varying positions, some of which are more proactive than others. The Medical Research Council, for example, requires academics to deposit all papers in a specific repository (UKPubMed Central) within six months of publication, but the Engineering and Physical Sciences Research Council has no such measures.
Now, after a study by SQW Consulting concluded that open access is increasingly popular with UK researchers and that institutions are setting up their own repositories, the councils will revisit the issue.
They will have to tread carefully because open access threatens to undermine the business model of publishers and learned societies.
In a statement, the councils said: "(we) have agreed that over time (we) will support increased open access by building on mandates on grant-holders to deposit research papers in suitable repositories within an agreed time period, and extending support for publishing in open-access journals, including through the pay-to-publish model."
The latest study lays out several options for the councils. These include sticking with the status quo, adopting a stronger common position favouring repositories, and backing the open-access journal model, which publishers could build into their existing business models.
The study also reports that more than three quarters of 2,100 council-funded researchers surveyed were unaware of the councils' current mandates.
Paul Gemmill, chair of the research outputs group at Research Councils UK, said the next stage was to decide whether a specific model should be adopted. He said the process would involve learned societies, publishers and academics.
Open-access advocate Stevan Harnad, professor of electronics and computer science at the University of Southampton, said scarce research money should not be used to pay open-access journal fees, where the costs normally borne by the publisher are picked up by funders.
"If good sense were to prevail, funders and universities would just mandate repositories," he said.
New guidance on who should pay for researchers to publish in open-access journals has also been issued by the Research Information Network and Universities UK. Arrangements for paying fees "remain haphazard", it says.