The public is being mislead by reports on teaching quality, the elite Russell Group of research-led universities has said.
In a confidential document, the Russell Group dismisses the results of the Quality Assurance Agency's £300 million-a-year teaching quality assessment exercise.
It says that many of the thousands of departmental reports, the key source of public information on quality and standards, are "unreliable and out of date".
The TQA, which has been running since 1992, will be replaced from the end of this year with a regime that all but abolishes subject-level inspection.
The Russell Group wants to initiate a campaign to prevent anyone compiling league tables from the published data.
The document says: "If subject-level review is discontinued, the existing scores, many of which are unreliable and out of date to the point of being misleading to potential students, will quickly become redundant."
"Every effort should be made to persuade newspapers and others not to use these data. Institutions could help this process by not using them in their own publicity."
The criticism comes from a group of universities that traditionally outperform all others in TQAs and league tables.
A series of data analyses by The THES , published during 2000 and 2001, has revealed serious questions about grade inflation, with the proportion of institutions gaining the highest TQA grades rising from about 25 per cent to more than 60 per cent, and about gamesmanship.
The Russell Group document, The QAA: A Statement of Principles and Proposals , was written last month and sent to funding chiefs and the QAA. It makes clear the extent to which the QAA caved in to the group's demands when agreeing a new quality assurance blueprint, published for consultation last week.
In words that echo the QAA's consultation document, the Russell Group document says that responsibility for standards and quality "should remain the responsibility of the institutions themselves".
"The national model for quality assurance should be one essentially of institutional review, not the current joint model of institutional and subject review."
The document criticises much of the work the QAA has done to develop a new regime since its launch two years ago.
"There are merits to some aspects of the quality infrastructure which the QAA has developed over the past two years (including the qualifications framework, subject benchmark statements and programme specifications)," it says.
"However, we feel that its components would have the status of guidance to institutions, with institutions deciding whether to use them or not. Specifically, the initial institutional profiles the QAA has been developing... should be discarded."