The sector is divided over whether some students should be allowed into university with lower entry grades than others because of their educational background, a government review of admissions suggested this week.
Asked whether some applicants should be made lower offers given the type of school they have attended, 38 per cent of admissions managers said "yes" and 38 per cent said "no", while 25 per cent were undecided.
Some universities make lower entry offers to students who have come from badly performing schools.
One manager argued: "If university policy is to achieve diversity, then policy and practice should reflect this aim. In these circumstances, differential selection is as justifiable as selection using any other criteria for entry."
But another commented: "Social engineering is not the purpose of the university."
Another opponent said: "The potential to achieve is the overriding consideration, and applicants should not be 'set up to fail' in order to satisfy generalised ideas that everyone should be given places in higher education institutions."
The survey of 160 university and college senior admissions managers was conducted as part of a review of how processes have changed since the 2004 Schwartz report on fair admissions to higher education.
It found that across the sector there was greater transparency over admissions decision-making, but that there were still some significant areas of concern, particularly around guidance for students who take vocational qualifications.
The report says: "The publicising of information about A levels was much more transparent than for other types of qualifications. Of those who stated that they prefer not to accept other vocational ... qualifications, only half publicise this on the institutional website." Admissions policies were found on 61 per cent of higher education institutions' websites.
More training was being provided to staff handling admissions, and there was evidence of movement towards more centralised admissions systems.
Among higher education institutions, 23 per cent were fully centralised, 20 per cent decentralised and 58 per cent a mixture.
The majority of respondents - 60 per cent - did not believe universities should choose students partly to achieve a social mix.
Forty-one per cent of universities and colleges thought an applicant's educational background, such as the type of school they attended, should be considered, a fall from 65 per cent in 2004.
The authors of the report, from Sheffield Hallam and Staffordshire universities, say that staff seem to be placing more weight on the actual academic attainment of applicants rather than potential achievement and conclude that this may mean they are being more "risk averse".
One respondent said: "We would welcome further discussion and consensus in the sector about the appropriateness and means of using contextual data in admissions decisions. The complexities that were outlined in the Schwartz report regarding the difficulties of assessing disadvantage, its impact upon an individual applicant's academic attainment, and its relevance to their potential to succeed in higher education are still unresolved."
Janet Graham, director of the Supporting Professionalism in Admissions Programme, said: "Things have improved since the Schwartz report, but there is still a way to go in terms of transparency ... awarding a mark out of 10 it is probably 7 or 8 but 'could do better '."