Schools which admit many pupils with special educational needs do not have lower attainment as a result, say researchers.
Schools face hard choices with pupils with behaviour problems
A study done for the Department for Education and Skills found "a very small and negative" relationship.
But there was considerable variation in the performance of schools with similar levels of special needs (SEN) pupils.
The likely reason for lower attainment was that highly "inclusive" schools served more disadvantaged communities.
The background to the study was a commitment by governments all over the world to greater inclusion in mainstream education of pupils with difficulties.
A recent report from England's education inspectorate, Ofsted, found schools wary of this, especially when pupils had behavioural problems.
The new report was compiled by a team from the universities of Manchester and Newcastle, and published on the Department for Education and Skills website.
They said schools were concerned their performance might be damaged - either in reality or in the way it was reported publicly - if they were to become "too" inclusive.
"This concern is also evident in the correspondence pages of the educational press and the statements of some teacher union representatives," they said.
So they tried to find whether schools did worse by many or all of their pupils because the presence of many pupils with SEN somehow distorted their processes - or did better, because they were skilful at responding to individual differences.
They made clear they had not investigated whether pupils with higher levels of SEN learnt more or less in mainstream settings than in special schools, or issues of rights, entitlement, resources or the current quality of SEN provision.
The researchers analysed primary and secondary school data, studied 16 highly inclusive schools and reviewed national and international reports on the subject.
In assessing pupils' attainment, they did not use the average figures which appear in school performance tables.
Inevitably, they said, the more low-attaining pupils in the school average, the lower the average would be - important in how performance was assessed and how the results were made public, but not about whether children could learn effectively in inclusive schools.
In secondary schools, they found the negative effects were stronger on pupils without SEN than those with.
In primary schools it was the other way round - pupils with SEN faring worse.
The relationship was more marked for schools with relatively few SEN pupils.
"Given these variations, it seems unlikely that the relationship is causal," they concluded.
Attainment was largely independent of levels of inclusivity.
Other factors - such as socio-economic status, gender, ethnicity and mother tongue - seemed "much more significant".
The implications were that schools need not feel anxious about becoming more inclusive, but should monitor the effects with care.
There was some evidence from teachers and pupils that inclusion could have positive effects on such things as social skills and understanding - especially for the SEN pupils.
Although not a model of "good practice" as such, highly inclusive schools had common features which meant they did not compromise pupil achievement - the key being flexibility.
"Pupils are neither rigidly segregated from their peers nor 'dumped' in mainstream classes, but are offered careful mixtures of provision in a range of settings."
The report suggested more work needed to be done, not least on the impact of different types of special educational need.
Inclusion and Pupil Achievement by Alan Dyson, Peter Farrell, Filiz Polat and Graeme Hutcheson, University of Manchester and Frances Gallannaugh, University of Newcastle.