Girls' steady academic improvement relative to boys has been the subject of much discussion since the early 1980s.
The Institute study concludes: "Social class and prior attainment remain the most powerful predictors of educational achievement."
The review was conducted by Jannette Elwood, a Qualifications and Curriculum Authority project manager for key stage 3 tests and Caroline Gipps, deputy vice-chancellor of Kingston University in London.
It draws on previous work by Professor Alan Smithers and Dr Pamela Robinson, which shows that the difference between single-sex and co-ed school performance among girls depends on the type of pupils they enrol.
In 1995 45.4 per cent of girls at mixed comprehensives gained GCSE grades A-C, compared with 45.5 per cent of girls at single-sex comprehensives.
Yet in the grant-maintained sector - where there is some selection - the figure increases from 50.8 in co-ed schools to 71.3 in girl-only ones, and in independent schools from 81.3 to 91.3.
The findings come against a background of increased demand for, and declining availability of, places at girls' schools.
Girls are often placed in single-sex education by parents who believe they are less likely to be distracted and will gain in confidence by being protected from the macho culture associated with science subjects and maths in co educational schools.
There is a bigger difference in achievement between the state and independent sectors than there is between co-ed and single-sex schools.
"Review of recent research on the achievement of girls in single-sex schools" by Janette Elwood and Caroline Gipps, is available from The Education Bookshop, Institute of Education, University of London, 20 Bedford Way, London WC1H 0AL, pound;6.95